Models

Every large language model we track: who makes it, what it costs through the API, how big its memory is, and whether it is that lab’s newest release. Prices re-verify daily; a dash means we hold no verified figure and will not guess.

ModelLabAPI $/1M in · outContextCommunity EloBattlesSentiment
Qwen3.7 Maxlatestqwen$1.475 · $4.4251000k14741W–0LPositive · 3
Grok 4.5latestx-ai$2 · $6500k14522W–2LPositive · 4
DeepSeek V4 Prolatestdeepseek$0.3969 · $0.79381049k14510W–1LPositive · 7
Kimi K3latestmoonshotai$3 · $151049k14422W–0LPositive · 20
DeepSeek V4 Flashlatestdeepseek$0.0489 · $0.09771049k14320W–1LPositive · 16
Claude Sonnet 5latestanthropic$2 · $101000kMixed · 5
GLM 5.2latestz-ai$0.966 · $3.0361049k2W–2LPositive · 9
GPT-5.6 Lunalatestopenai$0.2 · $1.21050kMixed · 12
GPT-5.6 Sollatestopenai$2 · $101050kMixed · 22
GPT-5.6 Terralatestopenai$2 · $121050k1W–3LNegative · 5
Mistral Medium 3.5latestmistralai$1.5 · $7.5262kMixed · 4
Nano Banana 2 (Gemini 3.1 Flash Image)latestgoogle$0.5 · $3131kPositive · 5
Claude Opus 4.6anthropic$5 · $251000k14970W–4LMixed · 4
Claude Fable 5anthropic$10 · $501000k1495Negative · 23
Gemini 3.5 Flashgoogle$1.5 · $91049k14833W–0LMixed · 16
Gemini 3.1 Pro Previewgoogle$2 · $121049k1480Negative · 11
GPT-5.5openai$5 · $301050k1466Negative · 13
Claude Opus 4.8anthropic$5 · $251000k1452Mixed · 12
Gemini 3.1 Flash Litegoogle$0.25 · $1.51049kMixed · 12
GPT-5.3-Codexopenai$1.75 · $14400kNegative · 6
GPT-5.4 Image 2openai$8 · $15272k

“latest” = that lab’s newest release among the models we track, computed from release dates, never asserted. API prices are OpenRouter’s blended figure and can move during the day. Community Elo is the LMArena public leaderboard. Sentiment is classified from public posts over the last 90 days; the number is how many carried an opinion.