What a working coding answer costs
Cost per solved task against coding index
Left is cheaper, up is better. Each dot is one model with both numbers.
70 of 356 models have both numbers. Cost per solved task is on a log scale, so each gridline is two to two and a half times the one before.
Coding index, 0 to 100 Cost per solved task, USD, log scale. Cheaper is left.
- OpenAI
- Qwen
- Mistral AI
- DeepSeek
- Anthropic
- MiniMax
- Meta
- NVIDIA
- Xiaomi
- inclusionAI
- Z.ai, Moonshot AI, xAI
- Other makers
- Faded: 1 benchmark. Solid: 2 or more.
- Frontier: the best index available at or below each cost
Cheap is left, good is up. DeepSeek V4 Flash 0731, GLM 5.3 Flash, GPT-5.6 Luna, GLM 5.3, GPT-5.6 Sol, Qwen3.8 27B and Claude Opus 5 sit on the frontier: nothing else is both cheaper and higher. Paying more does not move a model up: nothing that costs more than $1.33, the price of the highest score (83.0), beats 79.3. The dearest point, Mistral Medium 3.5 at $6.02, scores 25.5.
| Model | Maker | Cost per solved task | Index | Benchmarks |
|---|---|---|---|---|
| DeepSeek V4 Flash 0731 | DeepSeek | $0.01 | 51.8 | 4 |
| MiMo-V2.5 | Xiaomi | $0.01 | 50.0 | 3 |
| GLM 5.3 Flash | Z.ai | $0.02 | 57.7 | 6 |
| MiMo-V2.5-Pro | Xiaomi | $0.03 | 31.6 | 4 |
| GPT-5.6 Luna | OpenAI | $0.04 | 58.0 | 6 |
| Ling 3.0 Flash | inclusionAI | $0.05 | 29.6 | 2 |
| Nemotron 3.5 Lightning | NVIDIA | $0.06 | 22.6 | 2 |
| GPT-5 Mini | OpenAI | $0.08 | 34.2 | 2 |
| DeepSeek V4 Pro 0813 | DeepSeek | $0.10 | 56.1 | 5 |
| GPT-5.4 Nano | OpenAI | $0.14 | 44.8 | 2 |
| Inkling Small | Thinking Machines | $0.16 | 27.3 | 4 |
| Gemini 3.1 Flash Lite Preview | $0.18 | 22.7 | 3 | |
| Kimi K2.5 | Moonshot AI | $0.24 | 38.6 | 3 |
| GLM 4.7 | Z.ai | $0.29 | 55.1 | 2 |
| Grok 4.3 | xAI | $0.31 | 37.4 | 3 |
| Kimi K2.6 | Moonshot AI | $0.35 | 52.1 | 3 |
| Kimi K2.7 Code | Moonshot AI | $0.35 | 53.4 | 4 |
| GLM 5.3 | Z.ai | $0.36 | 70.6 | 6 |
| Gemini 3 Flash Preview | $0.37 | 43.7 | 2 | |
| GPT-5 | OpenAI | $0.39 | 38.8 | 3 |
| Qwen3.5-Flash | Qwen | $0.40 | 41.1 | 2 |
| Gemini 3.5 Flash Lite | $0.41 | 49.0 | 3 | |
| GPT-5.6 Terra | OpenAI | $0.42 | 63.9 | 6 |
| Muse Spark 1.1 | Meta | $0.43 | 70.3 | 3 |
| GPT-5.2-Codex | OpenAI | $0.43 | 45.4 | 3 |
| Claude Haiku 4.5 | Anthropic | $0.56 | 35.6 | 4 |
| MiniMax M3 | MiniMax | $0.56 | 39.2 | 6 |
| DeepSeek V4 Pro 0423 | DeepSeek | $0.57 | 38.7 | 5 |
| GPT-5.3-Codex | OpenAI | $0.59 | 60.0 | 3 |
| GLM 5.1 | Z.ai | $0.60 | 49.2 | 3 |
| MiniMax M2.5 | MiniMax | $0.62 | 37.2 | 3 |
| Grok 4.5 | xAI | $0.62 | 58.6 | 6 |
| GLM 5 | Z.ai | $0.63 | 41.7 | 3 |
| Gemini 2.5 Pro | $0.64 | 19.1 | 3 | |
| Muse Spark 1.2 | Meta | $0.64 | 55.7 | 5 |
| MiniMax M2.7 | MiniMax | $0.64 | 36.2 | 3 |
| GPT-5.4 Mini | OpenAI | $0.70 | 35.8 | 5 |
| Grok 4.20 | xAI | $0.72 | 33.3 | 2 |
| Inkling | Thinking Machines | $0.76 | 31.5 | 6 |
| Kimi K3 | Moonshot AI | $0.81 | 62.3 | 6 |
| Grok 4.6 | xAI | $0.82 | 65.7 | 6 |
| GLM 5.2 | Z.ai | $0.86 | 61.1 | 4 |
| GPT-5.1-Codex | OpenAI | $0.892 | 15.3 | 2 |
| DeepSeek V3.2 | DeepSeek | $0.99 | 30.0 | 3 |
| Gemini 3.1 Pro Preview | $0.99 | 32.3 | 4 | |
| GPT-5.1 | OpenAI | $1.02 | 40.4 | 3 |
| GPT-5.4 | OpenAI | $1.02 | 64.4 | 3 |
| Claude Sonnet 4.5 | Anthropic | $1.16 | 39.5 | 3 |
| Qwen3.6 Plus | Qwen | $1.19 | 44.1 | 3 |
| GPT-5.6 Sol | OpenAI | $1.20 | 73.5 | 6 |
| Gemini 3.5 Flash | $1.21 | 39.8 | 4 | |
| Qwen3.8 27B | Qwen | $1.26 | 74.2 | 4 |
| Qwen3.6 27B | Qwen | $1.30 | 36.8 | 2 |
| Qwen3.8 Max (0902) | Qwen | $1.32 | 62.4 | 5 |
| Claude Opus 5 | Anthropic | $1.33 | 83.0 | 6 |
| Qwen3.5 Plus 2026-04-20 | Qwen | $1.35 | 39.5 | 2 |
| Claude Opus 4.5 | Anthropic | $1.37 | 44.0 | 3 |
| Gemini 3.6 Flash | $1.49 | 62.5 | 4 | |
| Claude Opus 4.6 | Anthropic | $1.56 | 57.2 | 4 |
| GPT-5.5 | OpenAI | $1.65 | 71.3 | 5 |
| Claude Sonnet 4.6 | Anthropic | $1.68 | 57.3 | 5 |
| Gemini 3.7 Flash | $1.78 | 58.0 | 6 | |
| Claude Sonnet 5 | Anthropic | $1.87 | 57.7 | 6 |
| GPT-5.2 | OpenAI | $1.91 | 55.9 | 3 |
| Claude Fable 5 | Anthropic | $2.16 | 79.3 | 6 |
| Claude Opus 4.8 | Anthropic | $2.17 | 65.0 | 6 |
| Gemini 3.8 Flash | $2.74 | 61.6 | 6 | |
| Claude Opus 4.7 | Anthropic | $2.95 | 67.5 | 4 |
| Qwen3.7 Max | Qwen | $4.69 | 54.1 | 3 |
| Mistral Medium 3.5 | Mistral AI | $6.02 | 25.5 | 3 |
Source: OpenRouter, Aider polyglot leaderboard, LiveBench, SWE-bench (swebench.com), LMArena WebDev, Vals AI boards. Fetched October 22, 2026.