Cheapest coding LLMs
The cheapest models that hold up on SWE-bench, HumanEval, MBPP, LiveCodeBench. Ranked by price per 1M input tokens.
Ranked by input price
Models with meaningful coding benchmark scores, cheapest first.
| # | Model | In $/1M | Out $/1M | Type |
|---|---|---|---|---|
| 1 | $0.00 | $0.00 | OSS | |
| 2 | $0.00 | $0.00 | OSS | |
| 3 | $0.00 | $0.00 | Closed | |
| 4 | $0.00 | $0.00 | OSS | |
| 5 | $0.02 | $0.64 | Closed | |
| 6 | $0.02 | $0.09 | OSS | |
| 7 | $0.02 | $7.00 | OSS | |
| 8 | $0.03 | $4.80 | Closed | |
| 9 | $0.04 | $0.17 | OSS | |
| 10 | $0.04 | $0.15 | Closed | |
| 11 | $0.05 | $0.19 | OSS | |
| 12 | $0.05 | $0.40 | Closed | |
| 13 | $0.05 | $0.20 | Closed | |
| 14 | $0.06 | $0.25 | Closed | |
| 15 | $0.07 | $0.23 | OSS | |
| 16 | $0.07 | $0.28 | OSS | |
| 17 | $0.07 | $0.20 | Closed | |
| 18 | $0.08 | $0.45 | OSS | |
| 19 | $0.08 | $0.45 | OSS | |
| 20 | $0.08 | $0.28 | OSS | |
| 21 | $0.09 | $0.34 | OSS | |
| 22 | $0.09 | $0.29 | OSS | |
| 23 | $0.09 | $0.55 | OSS | |
| 24 | $0.09 | $1.10 | OSS | |
| 25 | $0.09 | $0.36 | Closed | |
| 26 | $0.10 | $0.50 | Closed | |
| 27 | $0.10 | $0.50 | Closed | |
| 28 | $0.10 | $0.30 | OSS | |
| 29 | $0.10 | $0.15 | OSS | |
| 30 | $0.10 | $0.30 | OSS |
Top 3 cheapest coding LLMs
gpt-oss-120b (free) is the cheapest model that clears our coding benchmark filter. At $0.00 per 1M input tokens, it lands well below frontier models while keeping credible SWE-bench and HumanEval scores.
gpt-oss-20b (free) follows close behind on price while holding up on coding-specific evals. Good fallback if gpt-oss-120b (free) is down or rate limited.
LFM2.5-2.6B (free) follows close behind on price while holding up on coding-specific evals. Good fallback if gpt-oss-120b (free) is down or rate limited.
The price gap · cheapest vs most expensive
Claude Haiku 5.5 is roughly Infinityx the price of gpt-oss-120b (free). The premium pays for long-context agentic coding, higher SWE-bench Verified scores, and vendor reliability. For one-shot code gen or refactors, the cheap options deliver 80 to 95 percent of the quality.