modelbenchmark.io

Price against quality

The frontier

Every scored model, plotted against what it costs to run. The models on the step are the ones nothing else beats on both price and score. Read it left to right: the cheapest model that clears the quality you need is the one you want.

Scope
Evidence

Overall percentile against output price

73 models plotted · 8 on the frontier · firm evidence only

  • On the frontier
  • Beaten on both
  • Open weights
020406080100$0.10$1.00$10.00$100.00
Output price, $ per million tokens · log scale →Overall percentile

Cheapest on the frontier

GLM-5.3-Flash

61st · $0.025 per million out

Best score at any price

GPT-6 Astra

100th · $50.00 per million out

On the frontier

ModelLabCompositeOut $/MIn $/MEvidence
GPT-6 AstraOpenAI100th$50.00$10.0011 bench · 2 src
Claude Opus 5Anthropic96th$25.00$5.009 bench · 2 src
Gemini 3.8 FlashGoogle95th$3.75$0.758 bench · 2 src
Gemini 3 Flash PreviewGoogle84th$3.00$0.5010 bench · 2 src
DeepSeek V4 Flash 0731DeepSeek81st$0.50$0.208 bench · 2 src
GPT-5.6 LunaOpenAI80th$0.37$0.068 bench · 2 src
DeepSeek-V3.2-ExpDeepSeek68th$0.326$0.21747 bench · 3 src
GLM-5.3-FlashZhipu61st$0.025$0.0757 bench · 2 src

Price is the catalog list price for output tokens, from the model's own vendor where it has a listing. A model with no price in the catalog cannot be plotted. See the full ranking.