Living index
LLM Pricing & Capability Index
One place to compare the major models on the things that decide it: price, context, size, and whether you can own the weights. Kept current and openly downloadable, so you (and your tools) can rely on it.
| Model | Maker | Weights | Params (total/active) | Context | $/M in | $/M out | |
|---|---|---|---|---|---|---|---|
| DeepSeek V3.x | DeepSeek | open | 671B / 37B | 128K | $0.20 | $0.80 | Try |
| Qwen3 (large) | Alibaba | open | 235B / 22B | 256K | $0.40 | $1.60 | |
| Llama 4 (large) | Meta | open | 400B / 17B | 256K | $0.50 | $1.60 | |
| Kimi K2.6 | Moonshot | open | 1000B / 32B | 256K | $0.60 | $2.50 | |
| GLM-5.2 | Z.ai | open | 744B / 40B | 1M | $0.91 | $2.86 | Try |
| Gemini (frontier) | closed | — | 1M | $2.00 | $12.00 | ||
| GPT-5.5 (class) | OpenAI | closed | — | 256K | $3.00 | $15.00 | |
| Claude (frontier) | Anthropic | closed | — | 200K | $3.00 | $15.00 |
Approximate public figures as of mid-2026; prices change often. Verify live provider pricing before relying on these numbers.
Changelog
- 2026-06-20Index published. Seeded with 8 frontier and open-weights models; GLM-5.2 added at launch.
Spotted a stale number? Tell me and it gets fixed and dated here.