The Best Value LLMs per Dollar
Which model buys the most benchmark quality per dollar?
18 models clear the 70-point floor and carry a metered price. The floor is an editorial line: without it the ratio always crowns the cheapest weak model. The cheapest of them outright is Nemotron 3 Ultra 550B at $1.00 per 1M tokens blended.
Method: Quality score divided by the blended price per 1M tokens (3:1 input-to-output tokens: 0.75 x input + 0.25 x output), among models scoring at least 70. Models priced at $0 (open weights with no metered API) are left out: you cannot divide by a price that does not exist. Pricing as of July 2026. 4 rows below carry a score measured at a non-default reasoning tier and are marked as such; every affected model is named on the hub. Read the full method.
| Rank | Model | Quality per $/1M | Quality | Blended $/1M | Evidence |
|---|---|---|---|---|---|
| #1 | Nex-N2-ProNex AGI | 75.60 | 75.6 | $1.00 | Both suites |
| #2 | Nemotron 3 Ultra 550BNVIDIA | 71.90 | 71.9 | $1.00 | SWE-bench only |
| #3 | MiMo-V2-ProXiaomi | 52.00 | 78.0 | $1.50 | SWE-bench only |
| #4 | Muse Spark 1.1Meta | 38.95 | 77.9 | $2.00 | Terminal-Bench only |
| #5 | GLM-5.2Zhipu· non-default tier | 36.23 | 77.9 | $2.15 | Terminal-Bench only |
| #6 | Grok 4.5xAI | 27.20 | 81.6 | $3.00 | Terminal-Bench only |
| #7 | Gemini 3.6 FlashGoogle | 25.83 | 77.5 | $3.00 | Terminal-Bench only |
| #8 | Gemini 3.5 FlashGoogle | 22.58 | 76.2 | $3.38 | Terminal-Bench only |
| #9 | Claude Sonnet 5Anthropic· non-default tier | 20.83 | 83.3 | $4.00 | Both suites |
| #10 | Qwen3.7 MaxAlibaba | 19.87 | 74.5 | $3.75 | Terminal-Bench only |
| #11 | Kimi K3Moonshot | 14.17 | 85.0 | $6.00 | Terminal-Bench only |
| #12 | GPT-5.6 TerraOpenAI | 12.85 | 72.3 | $5.63 | Terminal-Bench only |
| #13 | Claude Sonnet 4.6Anthropic· non-default tier | 11.87 | 71.2 | $6.00 | Terminal-Bench only |
| #14 | Claude Opus 4.8Anthropic | 8.70 | 87.0 | $10.00 | Both suites |
| #15 | Claude Opus 4.7Anthropic· non-default tier | 8.31 | 83.1 | $10.00 | Terminal-Bench only |
| #16 | GPT-5.6 SolOpenAI | 7.65 | 86.1 | $11.25 | Terminal-Bench only |
| #17 | GPT-5.5OpenAI | 7.16 | 80.5 | $11.25 | Terminal-Bench only |
| #18 | Claude Fable 5Anthropic | 4.23 | 84.6 | $20.00 | Terminal-Bench only |