PRICE INDEX · 60 MODELS · 3 SOURCES
LLM API Price Index
Straitly is the cheapest public source for all 36 Claude, GPT, Gemini and Grok models in this index, a median 40% below each maker's list price. Across 60 models from 9 makers, priced at a 3:1 input-to-output mix on 9 November 2026, Straitly is cheapest on 56 of 60.[1][3][7-15]
Last updated
What does one model cost on each source?
Pick a family and a model: official list, OpenRouter with its 5.5% fee, and Straitly, blended at 3:1 per 1M tokens.[1][3][4]
“OpenRouter's fee is charged when you buy credits, 5.5% on Standard and 8% on Business, never on individual requests.”
Anthropic API
$20.00/ 1M blended
The maker's list price, pay as you go.
- Input
- $10.00
- Output
- $50.00
- Cached input
- $0.25
OpenRouter
$21.10/ 1M blended
Includes OpenRouter's 5.5% credit fee. The rows below are its listed price.
- Input
- $10.00
- Output
- $50.00
- Cached input
- $0.25
Straitly
CHEAPEST$12.00/ 1M blended
40% below the list price, nothing added on top.
- Input
- $6.00
- Output
- $30.00
- Cached input
- $0.15
What is the cheapest API in each model family?
The lowest-priced model in each family and where it costs least, blended at 3:1 with OpenRouter's fee included.[6]
| FAMILY | LOWEST-PRICED MODEL | CHEAPEST SOURCE | BLENDED / 1M | OFFICIAL BLENDED | STRAITLY CHEAPEST |
|---|---|---|---|---|---|
| ClaudeAnthropic | Claude Haiku 5.5 | cheapest on Straitly | $0.12 /1M | list $0.20 | 15 of 15 Claude models cheapest on Straitly |
| GPTOpenAI | GPT-6 Luna | cheapest on Straitly | $0.12 /1M | list $0.20 | 6 of 6 GPT models cheapest on Straitly |
| GeminiGoogle | Gemini 2.5 Flash-Lite | cheapest on Straitly | $0.105 /1M | list $0.175 | 9 of 9 Gemini models cheapest on Straitly |
| DeepSeekDeepSeek | DeepSeek V4.1 Flash | cheapest on OpenRouter | $0.1366 /1M | list $0.2625 | 1 of 3 DeepSeek models cheapest on Straitly |
| KimiMoonshot AI | Kimi K2.5 | cheapest on Straitly | $0.54 /1M | list n/a | 5 of 5 Kimi models cheapest on Straitly |
| QwenAlibaba Cloud | Qwen3.7 Flash | cheapest on Straitly | $0.0291 /1M | list $0.055 | 7 of 7 Qwen models cheapest on Straitly |
| GLMZ.ai | GLM-5.3-Flash | cheapest on OpenRouter | $0.0484 /1M | list $0.2375 | 5 of 7 GLM models cheapest on Straitly |
| GrokxAI | Grok Build 0.1 | cheapest on Straitly | $0.75 /1M | list $1.25 | 6 of 6 Grok models cheapest on Straitly |
| MiniMaxMiniMax | MiniMax M3.1 Flash | cheapest on Straitly | $0.1496 /1M | list $0.2625 | 2 of 2 MiniMax models cheapest on Straitly |
Show all 9 rowsShow the first 6 rows
What does each LLM API cost per 1M tokens?
| OFFICIAL API | OPENROUTER | STRAITLY | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Claude Fable 5.1Anthropic | $10.00 | $50.00 | $0.25 | $10.00 | $50.00 | $0.25 | $6.00 | $30.00 | $0.15 | Straitly$12.00 blended |
| Claude Opus 5.5Anthropic | $4.00 | $20.00 | $0.20 | $4.00 | $20.00 | $0.20 | $2.40 | $12.00 | $0.12 | Straitly$4.80 blended |
| Claude Sonnet 5.5Anthropic | $2.00 | $10.00 | $0.10 | $2.00 | $10.00 | $0.10 | $1.20 | $6.00 | $0.12 | Straitly$2.40 blended |
| Claude Haiku 5.5[1]Anthropic | $0.10 | $0.50 | $0.01 | $0.10 | $0.50 | $0.01 | $0.06 | $0.30 | $0.006 | Straitly$0.12 blended |
| Claude Fable 5Anthropic | $10.00 | $50.00 | $1.00 | $10.00 | $50.00 | $1.00 | $6.00 | $30.00 | $0.60 | Straitly$12.00 blended |
| Claude Opus 5Anthropic | $5.00 | $25.00 | $0.50 | $5.00 | $25.00 | $0.50 | $3.00 | $15.00 | $0.30 | Straitly$6.00 blended |
| Claude Sonnet 5Anthropic | $2.00 | $10.00 | $0.20 | $2.00 | $10.00 | $0.20 | $1.20 | $6.00 | $0.12 | Straitly$2.40 blended |
| Claude Opus 4.8Anthropic | $5.00 | $25.00 | $0.50 | $5.00 | $25.00 | $0.50 | $3.00 | $15.00 | $0.30 | Straitly$6.00 blended |
| Claude Opus 4.7Anthropic | $5.00 | $25.00 | $0.50 | $5.00 | $25.00 | $0.50 | $3.00 | $15.00 | $0.30 | Straitly$6.00 blended |
| Claude Opus 4.6Anthropic | $5.00 | $25.00 | $0.50 | $5.00 | $25.00 | $0.50 | $3.00 | $15.00 | $0.30 | Straitly$6.00 blended |
| Claude Opus 4.5Anthropic | $5.00 | $25.00 | $0.50 | $5.00 | $25.00 | $0.50 | $3.00 | $15.00 | $0.30 | Straitly$6.00 blended |
| Claude Opus 4.1Anthropic | $15.00 | $75.00 | $1.50 | $15.00 | $75.00 | $1.50 | $9.00 | $45.00 | $0.90 | Straitly$18.00 blended |
| Claude Sonnet 4.6Anthropic | $3.00 | $15.00 | $0.30 | $3.00 | $15.00 | $0.30 | $1.80 | $9.00 | $0.18 | Straitly$3.60 blended |
| Claude Sonnet 4.5Anthropic | $3.00 | $15.00 | $0.30 | $3.00 | $15.00 | $0.30 | $1.80 | $9.00 | $0.18 | Straitly$3.60 blended |
| Claude Haiku 4.5Anthropic | $1.00 | $5.00 | $0.10 | $1.00 | $5.00 | $0.10 | $0.60 | $3.00 | $0.06 | Straitly$1.20 blended |
| GPT-6 AstraOpenAI | $10.00 | $50.00 | $1.00 | $10.00 | $50.00 | $1.00 | $6.00 | $30.00 | $0.60 | Straitly$12.00 blended |
| GPT-6.1 SolOpenAI | $2.00 | $10.00 | $0.10 | $2.00 | $10.00 | $0.10 | $1.20 | $6.00 | $0.06 | Straitly$2.40 blended |
| GPT-6.1 TerraOpenAI | $1.00 | $5.00 | $0.05 | $1.00 | $5.00 | $0.05 | $0.60 | $3.00 | $0.03 | Straitly$1.20 blended |
| GPT-6 LunaOpenAI | $0.10 | $0.50 | $0.01 | $0.10 | $0.50 | $0.01 | $0.06 | $0.30 | $0.006 | Straitly$0.12 blended |
| GPT-5.6 Sol[2]OpenAI | $4.00 | $20.00 | $0.40 | $4.00 | $20.00 | $0.40 | $1.70 | $8.50 | $0.17 | Straitly$3.40 blended |
| GPT-5.3 CodexOpenAI | $1.75 | $14.00 | $0.175 | $1.75 | $14.00 | $0.175 | $1.05 | $8.40 | $0.105 | Straitly$2.8875 blended |
| Gemini 4 Argon[3]Google | $2.00 | $10.00 | $0.10 | $2.00 | $10.00 | $0.10 | $1.20 | $6.00 | $0.06 | Straitly$2.40 blended |
| Gemini 3.8 Flash[4]Google | $0.75 | $3.75 | $0.075 | $0.825 | $4.125 | $0.0825 | $0.45 | $2.25 | $0.045 | Straitly$0.90 blended |
| Gemini 3.6 Flash[5]Google | $0.75 | $3.75 | $0.075 | $0.75 | $3.75 | $0.075 | $0.45 | $2.25 | $0.045 | Straitly$0.90 blended |
| Gemini 3.5 Flash-LiteGoogle | $0.30 | $2.50 | $0.03 | $0.30 | $2.50 | $0.03 | $0.18 | $1.50 | $0.018 | Straitly$0.51 blended |
| Gemini 3.1 Pro Preview[6]Google | $2.00 | $12.00 | $0.20 | $2.00 | $12.00 | $0.20 | $1.20 | $7.20 | $0.12 | Straitly$2.70 blended |
| Gemini 3.1 Flash-LiteGoogle | $0.25 | $1.50 | $0.025 | $0.25 | $1.50 | $0.025 | $0.15 | $0.90 | $0.015 | Straitly$0.3375 blended |
| Gemini 2.5 Pro[7]Google | $1.25 | $10.00 | $0.125 | $1.25 | $10.00 | $0.125 | $0.75 | $6.00 | $0.075 | Straitly$2.0625 blended |
| Gemini 2.5 FlashGoogle | $0.30 | $2.50 | $0.03 | $0.30 | $2.50 | $0.03 | $0.18 | $1.50 | $0.018 | Straitly$0.51 blended |
| Gemini 2.5 Flash-LiteGoogle | $0.10 | $0.40 | $0.01 | $0.10 | $0.40 | $0.01 | $0.06 | $0.24 | $0.006 | Straitly$0.105 blended |
| DeepSeek V4.1 Flash[8]DeepSeek | $0.15 | $0.60 | $0.003 | $0.006 | $0.50 | $0.005 | $0.09 | $0.36 | $0.0018 | OpenRouter$0.1366 blended |
| DeepSeek V4 Pro[9]DeepSeek | $0.66 | $1.98 | $0.022 | $0.111 | $4.44 | $0.097 | $0.792 | $2.376 | $0.0264 | Official API$0.99 blended |
| DeepSeek V3.2[10]DeepSeek | n/a | n/a | n/a | $0.26 | $0.38 | $0.13 | $0.156 | $0.228 | $0.078 | Straitly$0.174 blended |
| Kimi K3.1 ProMoonshot AI | $2.00 | $10.00 | $0.20 | $2.00 | $10.00 | $0.20 | $1.14 | $5.70 | $0.114 | Straitly$2.28 blended |
| Kimi K3Moonshot AI | $3.00 | $15.00 | $0.30 | $0.192 | $16.00 | $0.182 | $1.53 | $7.6488 | $0.1532 | Straitly$3.0597 blended |
| Kimi K2.7 Code[11]Moonshot AI | $0.95 | $4.00 | $0.19 | $0.60 | $3.50 | $0.16 | $0.408 | $2.04 | $0.0816 | Straitly$0.816 blended |
| Kimi K2.6[12]Moonshot AI | $0.95 | $4.00 | $0.16 | $0.65 | $3.41 | $0.15 | $0.36 | $1.83 | $0.078 | Straitly$0.7275 blended |
| Kimi K2.5[13]Moonshot AI | n/a | n/a | n/a | $0.532 | $3.325 | $0.209 | $0.27 | $1.35 | $0.042 | Straitly$0.54 blended |
| Qwen3.8 Max[14]Alibaba Cloud | $2.00 | $6.00 | n/a | $2.00 | $6.00 | $0.25 | $0.99 | $2.9706 | $0.1236 | Straitly$1.4852 blended |
| Qwen3.8 Flash[15]Alibaba Cloud | $0.15 | $0.47 | n/a | $0.15 | $0.47 | $0.016 | $0.0678 | $0.2292 | $0.0084 | Straitly$0.1082 blended |
| Qwen3.8 27B[16]Alibaba Cloud | $0.50 | $3.00 | n/a | $0.15 | $1.875 | $0.0375 | $0.1232 | $0.8799 | $0.0249 | Straitly$0.3124 blended |
| Qwen3.7 Max[17]Alibaba Cloud | $2.50 | $7.50 | n/a | $1.475 | $4.425 | $0.295 | $0.75 | $2.25 | $0.15 | Straitly$1.125 blended |
| Qwen3.7 Plus[18]Alibaba Cloud | $0.32 | $1.28 | n/a | $0.32 | $1.28 | $0.064 | $0.1656 | $0.6606 | $0.0336 | Straitly$0.2894 blended |
| Qwen3.7 Flash[19]Alibaba Cloud | $0.03 | $0.13 | n/a | $0.03 | $0.13 | $0.006 | $0.0168 | $0.066 | $0.0036 | Straitly$0.0291 blended |
| Qwen3.5 397B A17B[20]Alibaba Cloud | n/a | n/a | n/a | $0.60 | $3.60 | $0.12 | $0.1032 | $0.6192 | $0.1032 | Straitly$0.2322 blended |
| GLM-5.3Z.ai | $1.40 | $4.40 | $0.26 | $0.0699 | $1.40 | $0.0689 | $0.416 | $1.3866 | $0.0693 | OpenRouter$0.4246 blended |
| GLM-5.3-FlashZ.ai | $0.15 | $0.50 | $0.03 | $0.0335 | $0.083 | $0.03 | $0.0383 | $0.1275 | $0.0077 | OpenRouter$0.0484 blended |
| GLM-5.2Z.ai | $1.40 | $4.40 | $0.26 | $0.006 | $6.80 | $0.005 | $0.45 | $1.44 | $0.084 | Straitly$0.6975 blended |
| GLM-5.1Z.ai | $1.40 | $4.40 | $0.26 | $1.40 | $4.40 | $0.26 | $0.495 | $1.9806 | $0.099 | Straitly$0.8664 blended |
| GLM-5Z.ai | $1.00 | $3.20 | $0.20 | $1.00 | $3.20 | $0.20 | $0.60 | $1.92 | $0.12 | Straitly$0.93 blended |
| GLM-4.7Z.ai | $0.60 | $2.20 | $0.11 | $0.60 | $2.20 | $0.11 | $0.24 | $1.05 | $0.048 | Straitly$0.4425 blended |
| GLM-4.6Z.ai | $0.60 | $2.20 | $0.11 | $0.50 | $2.00 | $0.10 | $0.30 | $1.20 | $0.06 | Straitly$0.525 blended |
| Grok 4.7[21]xAI | $2.00 | $6.00 | $0.50 | $2.00 | $6.00 | $0.50 | $1.20 | $3.60 | $0.30 | Straitly$1.80 blended |
| Grok 4.6[22]xAI | $2.00 | $6.00 | $0.50 | $2.00 | $6.00 | $0.50 | $1.20 | $3.60 | $0.30 | Straitly$1.80 blended |
| Grok 4.5[23]xAI | $2.00 | $6.00 | $0.30 | $2.00 | $6.00 | $0.30 | $1.20 | $3.60 | $0.30 | Straitly$1.80 blended |
| Grok 4.20[24]xAI | $1.25 | $2.50 | $0.20 | $1.25 | $2.50 | $0.20 | $0.75 | $1.50 | $0.12 | Straitly$0.9375 blended |
| Grok 4.3[25]xAI | $1.25 | $2.50 | $0.20 | $1.25 | $2.50 | $0.20 | $0.75 | $1.50 | $0.12 | Straitly$0.9375 blended |
| Grok Build 0.1xAI | $1.00 | $2.00 | $0.20 | $1.00 | $2.00 | $0.20 | $0.60 | $1.20 | $0.12 | Straitly$0.75 blended |
| MiniMax M3[26]MiniMax | $0.30 | $1.20 | $0.06 | $0.30 | $1.20 | $0.06 | $0.168 | $0.66 | $0.0336 | Straitly$0.291 blended |
| MiniMax M3.1 FlashMiniMax | $0.15 | $0.60 | $0.015 | $0.15 | $0.60 | $0.015 | $0.0855 | $0.342 | $0.0086 | Straitly$0.1496 blended |
Price notes (26)
- [1] Claude Haiku 5.5: Anthropic's list price covers prompts up to 100K tokens. Longer prompts cost $0.50 input, $2.50 output and $0.05 cached read per 1M tokens.
- [2] GPT-5.6 Sol: OpenAI lists this as promotional pricing until 30 November 2026.
- [3] Gemini 4 Argon: Google's price runs through 30 November 2026. After that, Google lists $4 input, $20 output and $0.20 cached read per 1M tokens.
- [4] Gemini 3.8 Flash: Google's price runs through 31 December 2026. After that, Google lists $1.50 input and $7.50 output per 1M tokens.
- [5] Gemini 3.6 Flash: Google's price runs through 31 December 2026. After that, Google lists $1.50 input and $7.50 output per 1M tokens.
- [6] Gemini 3.1 Pro Preview: Google's price covers prompts up to 200K tokens. Longer prompts cost $4 input and $18 output per 1M tokens.
- [7] Gemini 2.5 Pro: Google's price covers prompts up to 200K tokens. Longer prompts cost $2.50 input and $15 output per 1M tokens.
- [8] DeepSeek V4.1 Flash: DeepSeek's official column is its off-peak price. During peak hours (01:00 to 04:00 and 06:00 to 10:00 UTC on weekdays) DeepSeek charges double. Prices are for the 1017 release. Peak price: $0.30 input, $1.20 output per 1M tokens.
- [9] DeepSeek V4 Pro: DeepSeek's official column is its off-peak price. During peak hours (01:00 to 04:00 and 06:00 to 10:00 UTC on weekdays) DeepSeek charges double. Prices are for the 0813 release. Peak price: $1.32 input, $3.96 output per 1M tokens.
- [10] DeepSeek V3.2: DeepSeek V3.2 is no longer on DeepSeek's official price list, so this page compares OpenRouter and Straitly only.
- [11] Kimi K2.7 Code: Moonshot AI closed the K2 series to new users on 31 October 2026 and plans to retire it by the end of December 2026.
- [12] Kimi K2.6: Moonshot AI closed the K2 series to new users on 31 October 2026 and plans to retire it by the end of December 2026.
- [13] Kimi K2.5: Kimi K2.5 is no longer on Moonshot AI's official price list.
- [14] Qwen3.8 Max: OpenRouter lists this model as the dated 0902 release. Alibaba Cloud's international price list shows no separate cached-read price for it.
- [15] Qwen3.8 Flash: Alibaba Cloud's international price list shows no separate cached-read price for this model.
- [16] Qwen3.8 27B: Alibaba Cloud's international price list shows no separate cached-read price for this model.
- [17] Qwen3.7 Max: Alibaba Cloud's international price list shows no separate cached-read price for this model.
- [18] Qwen3.7 Plus: Alibaba Cloud's price covers prompts up to 256K tokens and is a limited-time 20% discount on its $0.40 input and $1.60 output list.
- [19] Qwen3.7 Flash: Alibaba Cloud's price covers prompts up to 32K tokens. Longer prompts cost up to $0.20 input and $0.80 output per 1M tokens.
- [20] Qwen3.5 397B A17B: Qwen3.5 397B A17B is not on Alibaba Cloud's international price list.
- [21] Grok 4.7: xAI's price covers prompts under 200K tokens. Longer prompts cost twice as much.
- [22] Grok 4.6: xAI's price covers prompts under 200K tokens. Longer prompts cost twice as much.
- [23] Grok 4.5: xAI's price covers prompts under 200K tokens. Longer prompts cost twice as much.
- [24] Grok 4.20: xAI's price covers prompts under 200K tokens. Longer prompts cost twice as much.
- [25] Grok 4.3: xAI's price covers prompts under 200K tokens. Longer prompts cost twice as much.
- [26] MiniMax M3: MiniMax's price is its Standard tier for prompts up to 512K tokens, after a permanent 50% cut from its $0.60 input and $2.40 output list.
Sources, dates and 4 more figures
- Official:
- each maker's pricing page, checked 9 November 2026[7-15].
- Straitly:
- its live public catalog, fetched 2026-11-09 08:04 UTC[1].
Straitly's full catalog, a wider set, shares 141 models with OpenRouter; there Straitly is below the cheapest OpenRouter seller on 129 of 141 (snapshot 9 November 2026). The 12 where an OpenRouter seller is cheaper are all open-weight models: Qwen, GLM, Gemma, DeepSeek, Kimi and gpt-oss variants. See Straitly vs OpenRouter.[5]
How much can the same model cost?
The priciest source costs a median 1.8x the cheapest, so switching saves a median 43%. The widest gaps:[6]
- $0.2322 on Straitly vs $1.4243 on OpenRouter
- QWEN3.5 397B A17B
- 6.1x
- $0.4246 on OpenRouter vs $2.15 on Z.ai's API
- GLM-5.3
- 5.1x
- $0.0484 on OpenRouter vs $0.2375 on Z.ai's API
- GLM-5.3-FLASH
- 4.9x
- $0.3124 on Straitly vs $1.125 on Alibaba Cloud's API
- QWEN3.8 27B
- 3.6x
- $1.125 on Straitly vs $3.75 on Alibaba Cloud's API
- QWEN3.7 MAX
- 3.3x
- $0.6975 on Straitly vs $2.15 on Z.ai's API
- GLM-5.2
- 3.1x
- Claude (15)
- 40% below5% above
- GPT (6)
- 40% below5% above
- Gemini (9)
- 40% below5% above
- DeepSeek (2)
- 10% below10% below
- Kimi (4)
- 51% below18% below
- Qwen (6)
- 52% below5% above
- GLM (7)
- 60% below8% below
- Grok (6)
- 40% below5% above
- MiniMax (2)
- 44% below5% above
Median of each model's blended 3:1 price against its official list price. Bracket = models with an official price. Official prices checked 9 November 2026; OpenRouter and Straitly prices fetched 9 November 2026.
Spread table for all 60 models
| MODEL | CHEAPEST SOURCE | PRICIEST SOURCE | GAP / 1M | PRICIER BY | SAVING |
|---|---|---|---|---|---|
| Qwen3.5 397B A17B | Straitly $0.2322 | OpenRouter $1.4243 | $1.1921 | 6.1x | 84% |
| GLM-5.3 | OpenRouter $0.4246 | Z.ai's API $2.15 | $1.7254 | 5.1x | 80% |
| GLM-5.3-Flash | OpenRouter $0.0484 | Z.ai's API $0.2375 | $0.1891 | 4.9x | 80% |
| Qwen3.8 27B | Straitly $0.3124 | Alibaba Cloud's API $1.125 | $0.8126 | 3.6x | 72% |
| Qwen3.7 Max | Straitly $1.125 | Alibaba Cloud's API $3.75 | $2.625 | 3.3x | 70% |
| GLM-5.2 | Straitly $0.6975 | Z.ai's API $2.15 | $1.4525 | 3.1x | 68% |
| GLM-5.1 | Straitly $0.8664 | OpenRouter $2.2682 | $1.4018 | 2.6x | 62% |
| GPT-5.6 Sol | Straitly $3.40 | OpenRouter $8.44 | $5.04 | 2.5x | 60% |
| Kimi K2.5 | Straitly $0.54 | OpenRouter $1.2979 | $0.7579 | 2.4x | 58% |
| GLM-4.7 | Straitly $0.4425 | OpenRouter $1.055 | $0.6125 | 2.4x | 58% |
| Kimi K2.6 | Straitly $0.7275 | Moonshot AI's API $1.7125 | $0.985 | 2.4x | 58% |
| Qwen3.8 Flash | Straitly $0.1082 | OpenRouter $0.2426 | $0.1345 | 2.2x | 55% |
| Qwen3.8 Max | Straitly $1.4852 | OpenRouter $3.165 | $1.6799 | 2.1x | 53% |
| Kimi K2.7 Code | Straitly $0.816 | Moonshot AI's API $1.7125 | $0.8965 | 2.1x | 52% |
| Qwen3.7 Plus | Straitly $0.2894 | OpenRouter $0.5908 | $0.3015 | 2.0x | 51% |
| Qwen3.7 Flash | Straitly $0.0291 | OpenRouter $0.058 | $0.0289 | 2.0x | 50% |
| Kimi K3 | Straitly $3.0597 | Moonshot AI's API $6.00 | $2.9403 | 2.0x | 49% |
| Gemini 3.8 Flash | Straitly $0.90 | OpenRouter $1.7407 | $0.8407 | 1.9x | 48% |
| DeepSeek V4.1 Flash | OpenRouter $0.1366 | DeepSeek's API $0.2625 | $0.1259 | 1.9x | 48% |
| GLM-4.6 | Straitly $0.525 | Z.ai's API $1.00 | $0.475 | 1.9x | 48% |
| MiniMax M3 | Straitly $0.291 | OpenRouter $0.5539 | $0.2629 | 1.9x | 47% |
| Kimi K3.1 Pro | Straitly $2.28 | OpenRouter $4.22 | $1.94 | 1.9x | 46% |
| MiniMax M3.1 Flash | Straitly $0.1496 | OpenRouter $0.2769 | $0.1273 | 1.9x | 46% |
| Gemini 2.5 Flash-Lite | Straitly $0.105 | OpenRouter $0.1846 | $0.0796 | 1.8x | 43% |
| DeepSeek V3.2 | Straitly $0.174 | OpenRouter $0.306 | $0.132 | 1.8x | 43% |
| Grok 4.7 | Straitly $1.80 | OpenRouter $3.165 | $1.365 | 1.8x | 43% |
| Grok 4.6 | Straitly $1.80 | OpenRouter $3.165 | $1.365 | 1.8x | 43% |
| Grok 4.5 | Straitly $1.80 | OpenRouter $3.165 | $1.365 | 1.8x | 43% |
| Claude Opus 5.5 | Straitly $4.80 | OpenRouter $8.44 | $3.64 | 1.8x | 43% |
| Claude Sonnet 5.5 | Straitly $2.40 | OpenRouter $4.22 | $1.82 | 1.8x | 43% |
| Claude Haiku 5.5 | Straitly $0.12 | OpenRouter $0.211 | $0.091 | 1.8x | 43% |
| Claude Sonnet 5 | Straitly $2.40 | OpenRouter $4.22 | $1.82 | 1.8x | 43% |
| Claude Opus 4.1 | Straitly $18.00 | OpenRouter $31.65 | $13.65 | 1.8x | 43% |
| Claude Sonnet 4.6 | Straitly $3.60 | OpenRouter $6.33 | $2.73 | 1.8x | 43% |
| Claude Sonnet 4.5 | Straitly $3.60 | OpenRouter $6.33 | $2.73 | 1.8x | 43% |
| Claude Haiku 4.5 | Straitly $1.20 | OpenRouter $2.11 | $0.91 | 1.8x | 43% |
| GPT-6.1 Sol | Straitly $2.40 | OpenRouter $4.22 | $1.82 | 1.8x | 43% |
| GPT-6.1 Terra | Straitly $1.20 | OpenRouter $2.11 | $0.91 | 1.8x | 43% |
| GPT-6 Luna | Straitly $0.12 | OpenRouter $0.211 | $0.091 | 1.8x | 43% |
| GPT-5.3 Codex | Straitly $2.8875 | OpenRouter $5.0772 | $2.1897 | 1.8x | 43% |
| Gemini 4 Argon | Straitly $2.40 | OpenRouter $4.22 | $1.82 | 1.8x | 43% |
| Gemini 3.6 Flash | Straitly $0.90 | OpenRouter $1.5825 | $0.6825 | 1.8x | 43% |
| Gemini 2.5 Pro | Straitly $2.0625 | OpenRouter $3.6266 | $1.5641 | 1.8x | 43% |
| GLM-5 | Straitly $0.93 | OpenRouter $1.6352 | $0.7052 | 1.8x | 43% |
| Grok 4.20 | Straitly $0.9375 | OpenRouter $1.6484 | $0.7109 | 1.8x | 43% |
| Grok 4.3 | Straitly $0.9375 | OpenRouter $1.6484 | $0.7109 | 1.8x | 43% |
| Claude Fable 5.1 | Straitly $12.00 | OpenRouter $21.10 | $9.10 | 1.8x | 43% |
| Claude Fable 5 | Straitly $12.00 | OpenRouter $21.10 | $9.10 | 1.8x | 43% |
| Claude Opus 5 | Straitly $6.00 | OpenRouter $10.55 | $4.55 | 1.8x | 43% |
| Claude Opus 4.8 | Straitly $6.00 | OpenRouter $10.55 | $4.55 | 1.8x | 43% |
| Claude Opus 4.7 | Straitly $6.00 | OpenRouter $10.55 | $4.55 | 1.8x | 43% |
| Claude Opus 4.6 | Straitly $6.00 | OpenRouter $10.55 | $4.55 | 1.8x | 43% |
| Claude Opus 4.5 | Straitly $6.00 | OpenRouter $10.55 | $4.55 | 1.8x | 43% |
| GPT-6 Astra | Straitly $12.00 | OpenRouter $21.10 | $9.10 | 1.8x | 43% |
| Gemini 3.5 Flash-Lite | Straitly $0.51 | OpenRouter $0.8967 | $0.3867 | 1.8x | 43% |
| Gemini 3.1 Pro Preview | Straitly $2.70 | OpenRouter $4.7475 | $2.0475 | 1.8x | 43% |
| Gemini 3.1 Flash-Lite | Straitly $0.3375 | OpenRouter $0.5934 | $0.2559 | 1.8x | 43% |
| Gemini 2.5 Flash | Straitly $0.51 | OpenRouter $0.8967 | $0.3867 | 1.8x | 43% |
| Grok Build 0.1 | Straitly $0.75 | OpenRouter $1.3187 | $0.5687 | 1.8x | 43% |
| DeepSeek V4 Pro | DeepSeek's API $0.99 | OpenRouter $1.2589 | $0.2689 | 1.3x | 21% |
Spread = priciest source / cheapest source for the same model, each blended at 3 input tokens per output token with OpenRouter's 5.5% credit fee added. Saving = what you keep by buying at the cheapest source instead of the priciest. Official prices checked 9 November 2026; OpenRouter and Straitly prices fetched 9 November 2026.
How much does cached input save?
Re-reading a cached prompt prefix costs a median 90% less than fresh input on the official APIs. For agents and RAG, cache hit rate moves the bill as much as where you buy.[7-15]
“A cache hit costs 10 percent of the standard input price”
- 51 models with a cached price
- MEDIAN OFFICIAL CACHE DISCOUNT
- 90%
- DeepSeek V4.1 Flash, $0.003 vs $0.15
- DEEPEST CACHE DISCOUNT
- 98%
- Grok 4.7, $0.50 vs $2.00
- SHALLOWEST CACHE DISCOUNT
- 75%
- 59 models, vs Straitly input price
- MEDIAN STRAITLY CACHE DISCOUNT
- 90%
Cache discount by model family
| FAMILY | OFFICIAL MEDIAN | OFFICIAL RANGE | STRAITLY MEDIAN | CHEAPEST CACHED READ / 1M |
|---|---|---|---|---|
| Claude | 90% off | 90% to 98% (15) | 90% off | $0.006 Claude Haiku 5.5, Straitly |
| GPT | 90% off | 90% to 95% (6) | 90% off | $0.006 GPT-6 Luna, Straitly |
| Gemini | 90% off | 90% to 95% (9) | 90% off | $0.006 Gemini 2.5 Flash-Lite, Straitly |
| DeepSeek | 97% off | 97% to 98% (2) | 97% off | $0.0018 DeepSeek V4.1 Flash, Straitly |
| Kimi | 87% off | 80% to 90% (4) | 84% off | $0.042 Kimi K2.5, Straitly |
| Qwen | n/a | n/a | 80% off | $0.0036 Qwen3.7 Flash, Straitly |
| GLM | 81% off | 80% to 82% (7) | 80% off | $0.0053 GLM-5.2, OpenRouter |
| Grok | 82% off | 75% to 85% (6) | 78% off | $0.12 Grok 4.20, Straitly |
| MiniMax | 85% off | 80% to 90% (2) | 85% off | $0.0086 MiniMax M3.1 Flash, Straitly |
Discount = 1 minus cached-read price divided by input price, within the same source. Cache writes and storage are priced separately by some makers and are not included. Cheapest cached read includes OpenRouter's 5.5% fee. Official prices checked 9 November 2026; OpenRouter and Straitly prices fetched 9 November 2026.
How is the index built?
Three public sources per model, one blend, one fee, lowest wins. Straitly is the lowest on 56 of 60.
1Three public prices per model
Official from each maker's pricing page; OpenRouter and Straitly from their public APIs.[1][3][7-15]
2One 3:1 blend
(3 x input + 1 x output) / 4 gives one price per 1M tokens.
3OpenRouter's fee is added
Its 5.5% credit fee goes on top before the three are compared.[4]
4The lowest wins
A tie goes to the official API.[6]
Take the data with you
60 rows, 22 columns: input, output and cached-input prices for each source, with the source URL and the date it was read.
The 22 columns
model, family, maker, official_input_usd_per_1m, official_output_usd_per_1m, official_cached_input_usd_per_1m, official_source, official_checked, openrouter_model_id, openrouter_input_usd_per_1m, openrouter_output_usd_per_1m, openrouter_cached_input_usd_per_1m, openrouter_source, openrouter_fetched, straitly_model_id, straitly_input_usd_per_1m, straitly_output_usd_per_1m, straitly_cached_input_usd_per_1m, straitly_source, straitly_fetched, cheapest_at_3to1_incl_openrouter_fee, note
Method: six rules behind every number
- Official API. The standard pay-as-you-go price on the maker's own pricing page, read by hand on 9 November 2026. Where a maker shows peak and off-peak or short and long prompt prices, the index uses the lower standard price and says so in the price notes. Batch and priority tiers are left out.
- OpenRouter.The per-token price its public models API lists for the model id, times 1M. OpenRouter says: "Inference is billed at the provider's list price on every plan. OpenRouter's fee is charged when you buy credits, 5.5% on Standard and 8% on Business, never on individual requests." The table shows the listed price; the cheapest column adds 5.5%.[4]
- Straitly. The price in Straitly's live public catalog, the same data behind straitly.ai/models. Straitly bills prepaid, per token, with no subscription and no minimum (pricing).[1][2]
- Cheapest. (3 x input + 1 x output) / 4, the lowest wins and a tie goes to the official API. Cached-input prices are listed for reference but not blended. A mostly-input or mostly-output workload can change the winner, so each model page also prices three monthly workloads.
- Inclusion. A model is listed only when at least two of the three sources publish a price for it.
- Two denominators. Counts on this page (56 of 60) cover the 60 models in this index and OpenRouter's listed price for each model id. The compare pages use Straitly's full public catalog (164 models on 9 November 2026) against the cheapest OpenRouter seller per model, which is where 129 of 141comes from. Different sets: Straitly's own prices match where they overlap, but each set takes its reference prices from its own source, so a reference price can differ between them.[5]
Models cheaper on another source (4)
In this index, 3 are cheapest on OpenRouter and 1on the maker's own API.[5][6]
| MODEL | CHEAPEST | ITS PRICE | STRAITLY | STRAITLY ABOVE BY |
|---|---|---|---|---|
| DeepSeek V4.1 Flash | OpenRouter | $0.1366 | $0.1575 | 15% |
| DeepSeek V4 Pro | DeepSeek's API | $0.99 | $1.188 | 20% |
| GLM-5.3 | OpenRouter | $0.4246 | $0.6587 | 55% |
| GLM-5.3-Flash | OpenRouter | $0.0484 | $0.0606 | 25% |
Blended 3:1, OpenRouter fee included. Official prices checked 9 November 2026; OpenRouter and Straitly prices fetched 9 November 2026.
Price history: dated changes, standing cuts, generations
Makers do not publish full price histories, so this sticks to what their pricing pages state, checked 9 November 2026.[7-15]
Dated price changes
HAPPENED
Kimi K2.6: K2 series closed to new users. Moonshot AI plans to retire it by the end of December 2026. Source: Moonshot AI pricing.
SCHEDULED
Gemini 4 Argon: Promotional price ends. Google lists $4 input and $20 output per 1M tokens after this date, double today's $2 and $10. Source: Google pricing.
SCHEDULED
GPT-5.6 Sol: OpenAI's promotional pricing for GPT-5.6 Sol ends. Source: OpenAI pricing.
SCHEDULED
Gemini 3.8 Flash: Promotional price ends. Google lists $1.50 input and $7.50 output after this date, double today's $0.75 and $3.75. Source: Google pricing.
SCHEDULED
Gemini 3.6 Flash: Promotional price ends. Google lists $1.50 input and $7.50 output after this date. Source: Google pricing.
Standing list-price cuts
| MODEL | WAS (IN / OUT) | NOW (IN / OUT) | CHANGE | TYPE |
|---|---|---|---|---|
| MiniMax M3 | $0.60 / $2.40 | $0.30 / $1.20 | 50% lower | permanent 50% cut |
| Qwen3.7 Plus | $0.40 / $1.60 | $0.32 / $1.28 | 20% lower | limited-time 20% discount |
"Was" is the higher list price each maker still shows next to its discount, per 1M input / output tokens.
Price by generation
- official blended, $30.00 to $8.00
- CLAUDE OPUS 4.1 TO CLAUDE OPUS 5.5
- -73%
- official blended, $6.00 to $4.00
- CLAUDE SONNET 4.5 TO CLAUDE SONNET 5.5
- -33%
- official blended, $2.00 to $0.20
- CLAUDE HAIKU 4.5 TO CLAUDE HAIKU 5.5
- -90%
- official blended, $3.4375 to $4.00
- GEMINI 2.5 PRO TO GEMINI 4 ARGON
- +16%
- official blended, $1.00 to $2.15
- GLM-4.6 TO GLM-5.3
- +115%
- official blended, $1.7125 to $4.00
- KIMI K2.6 TO KIMI K3.1 PRO
- +134%
| MODEL | RELEASED | INPUT | OUTPUT | VS PREVIOUS |
|---|---|---|---|---|
| Claude Opus 4.1 | 5 August 2025 | $15.00 | $75.00 | baseline |
| Claude Opus 4.5 | 24 November 2025 | $5.00 | $25.00 | 67% cheaper |
| Claude Opus 4.6 | 4 February 2026 | $5.00 | $25.00 | same |
| Claude Opus 4.7 | 16 April 2026 | $5.00 | $25.00 | same |
| Claude Opus 4.8 | 27 May 2026 | $5.00 | $25.00 | same |
| Claude Opus 5 | 24 July 2026 | $5.00 | $25.00 | same |
| Claude Opus 5.5 | 22 September 2026 | $4.00 | $20.00 | 20% cheaper |
| MODEL | RELEASED | INPUT | OUTPUT | VS PREVIOUS |
|---|---|---|---|---|
| Claude Sonnet 4.5 | 29 September 2025 | $3.00 | $15.00 | baseline |
| Claude Sonnet 4.6 | 17 February 2026 | $3.00 | $15.00 | same |
| Claude Sonnet 5 | 30 June 2026 | $2.00 | $10.00 | 33% cheaper |
| Claude Sonnet 5.5 | 28 September 2026 | $2.00 | $10.00 | same |
| MODEL | RELEASED | INPUT | OUTPUT | VS PREVIOUS |
|---|---|---|---|---|
| Claude Haiku 4.5 | 15 October 2025 | $1.00 | $5.00 | baseline |
| Claude Haiku 5.5 | 7 October 2026 | $0.10 | $0.50 | 90% cheaper |
| MODEL | RELEASED | INPUT | OUTPUT | VS PREVIOUS |
|---|---|---|---|---|
| Gemini 2.5 Pro | 17 June 2025 | $1.25 | $10.00 | baseline |
| Gemini 3.1 Pro Preview | 19 February 2026 | $2.00 | $12.00 | 31% pricier |
| Gemini 4 Argon | 13 October 2026 | $2.00 | $10.00 | 11% cheaper |
| MODEL | RELEASED | INPUT | OUTPUT | VS PREVIOUS |
|---|---|---|---|---|
| GLM-4.6 | 30 September 2025 | $0.60 | $2.20 | baseline |
| GLM-4.7 | 22 December 2025 | $0.60 | $2.20 | same |
| GLM-5 | 11 February 2026 | $1.00 | $3.20 | 55% pricier |
| GLM-5.1 | 7 April 2026 | $1.40 | $4.40 | 39% pricier |
| GLM-5.2 | 16 June 2026 | $1.40 | $4.40 | same |
| GLM-5.3 | 18 August 2026 | $1.40 | $4.40 | same |
| MODEL | RELEASED | INPUT | OUTPUT | VS PREVIOUS |
|---|---|---|---|---|
| Kimi K2.6 | 20 April 2026 | $0.95 | $4.00 | baseline |
| Kimi K3 | 16 July 2026 | $3.00 | $15.00 | 250% pricier |
| Kimi K3.1 Pro | n/a | $2.00 | $10.00 | 33% cheaper |
Current official list price for each generation, read from the maker's pricing page on 9 November 2026. "Vs previous" compares the 3:1 blended price with the generation above it. Release dates come from Straitly's public model catalog, fetched 9 November 2026.
What do buyers ask about LLM API prices?
Is Straitly always the cheapest LLM API?
Not always. Straitly is the cheapest of the three sources on all 36 Claude, GPT, Gemini and Grok models and on 56 of the 60 models in this index. It is not the cheapest on DeepSeek V4.1 Flash (OpenRouter is 13% lower), DeepSeek V4 Pro (DeepSeek's API is 17% lower), GLM-5.3 (OpenRouter is 36% lower), GLM-5.3-Flash (OpenRouter is 20% lower). Straitly's full catalog, a wider set, shares 141 models with OpenRouter; there Straitly is below the cheapest OpenRouter seller on 129 of 141 (snapshot 9 November 2026). The 12 where an OpenRouter seller is cheaper are all open-weight models: Qwen, GLM, Gemma, DeepSeek, Kimi and gpt-oss variants.
What is the cheapest Claude API?
Straitly is the cheapest of the three sources on all 15 Claude models in the index at a 3:1 input-to-output mix. The lowest-priced Claude model is Claude Haiku 5.5: $0.06 input and $0.30 output per 1M tokens on Straitly, against $0.10 and $0.50 on Anthropic's API.
What is the cheapest DeepSeek API?
Straitly is the cheapest of the three sources on 1 of 3 DeepSeek models at a 3:1 input-to-output mix. OpenRouter is cheaper for DeepSeek V4.1 Flash and DeepSeek's own API is cheaper for DeepSeek V4 Pro. The lowest-priced DeepSeek model is DeepSeek V4.1 Flash: $0.006 input and $0.50 output per 1M tokens on OpenRouter, against $0.15 and $0.60 on DeepSeek's API.
What is the cheapest Kimi API?
Straitly is the cheapest of the three sources on all 5 Kimi models in the index at a 3:1 input-to-output mix. The lowest-priced Kimi model is Kimi K2.5: $0.27 input and $1.35 output per 1M tokens on Straitly.
What is the cheapest Qwen API?
Straitly is the cheapest of the three sources on all 7 Qwen models in the index at a 3:1 input-to-output mix. The lowest-priced Qwen model is Qwen3.7 Flash: $0.0168 input and $0.066 output per 1M tokens on Straitly, against $0.03 and $0.13 on Alibaba Cloud's API.
3 more questions: method, fees and price gaps
- How much does the same model cost from different providers?
- Across the 60 models, the priciest source costs a median 1.8x the cheapest one, blended at 3:1 with OpenRouter's 5.5% fee included. The widest gap is Qwen3.5 397B A17B: $0.2322 per 1M blended tokens on Straitly against $1.4243 on OpenRouter, 6.1x apart.
- Does OpenRouter charge more than the list price?
- OpenRouter's pricing page says: "Inference is billed at the provider's list price on every plan. OpenRouter's fee is charged when you buy credits, 5.5% on Standard and 8% on Business, never on individual requests." So on the Standard plan every dollar of usage costs about $1.055 of card spend. Its listed per-token price for some open-weight models is below the official price, and the index shows those cases.
- How is the index built, and can I reuse the data?
- Straitly and OpenRouter prices come from their live public APIs; official prices are read by hand on each maker's pricing page. For each model the index blends input and output at 3:1, adds OpenRouter's 5.5% credit fee, and picks the lowest; a tie goes to the official API. This edition is dated 9 November 2026. The full table is a free CSV at straitly.ai/llm-api-price-index.csv, with a source and date for every price.
Sources
Numbers in brackets link here. Every price also carries its own source URL and check date in the CSV.
- [1]Straitly live model catalog (straitly.ai/api/catalog)Straitly, fetched 2026-11-09 08:04 UTC
- [2]Straitly pricingStraitly, checked 9 November 2026
- [3]OpenRouter models API (openrouter.ai/api/v1/models)OpenRouter, fetched 2026-11-09 08:04 UTC
- [4]OpenRouter pricing (openrouter.ai/pricing)OpenRouter, checked 9 November 2026
- [5]Straitly vs OpenRouter: full-catalog price snapshotStraitly, snapshot 9 November 2026
- [6]LLM API Price Index, CSV downloadStraitly, dated 9 November 2026
- [7-15]Official API pricing pages, checked 9 November 2026: [7] Anthropic, [8] OpenAI, [9] Google, [10] DeepSeek, [11] Moonshot AI, [12] Alibaba Cloud, [13] Z.ai, [14] xAI, [15] MiniMax
One key for every row.
Every model in the Straitly column runs through one OpenAI-compatible endpoint. Prepaid, no subscription, no minimum.