Capability badges appear only where the provider catalog explicitly lists support. A blank cell means the source is silent, not that the feature is absent.
Listed rates
Price across providers
Input price per million tokens as published by each provider. Bars are drawn from listed rates only — no traffic weighting, since the catalog observes no requests.
Lowest input$0.9/M
Across 44 priced providers
Median input$3/M
Midpoint of listed rates
Highest input$3.6/M
4.0× the lowest listed rate
Output range$5.55 – $18
Per million output tokens
Xpersona$0.9/MLowest
Auriko$3/M
Ofox$3/M
Merge Gateway$3/M
Pioneer$3/M
Opper$3/M
Requesty$3/M
Azure Cognitive Services$3/M
Abacus$3/M
routing.run$3/M
Cloudflare AI Gateway$3/M
Azure$3/M
CrossModel$3/M
DigitalOcean$3/M
FreeModel$3/M
Neon$3/M
LLM Gateway$3/M
Eden AI$3/M
Kilo Gateway$3/M
FrogBot$3/M
Impossibl$3/M
Cloudflare AI Gateway$3/M
GitHub Copilot$3/M
Modelis$3/M
Anthropic$3/M
OpenRouter$3/M
Perplexity Agent$3/M
Databricks$3/M
NanoGPT$3/M
FastRouter$3/M
NEAR AI Cloud$3/M
DaoXE$3/M
DevPass (LLM Gateway)$3/M
OpenCode Zen$3/M
Vercel AI Gateway$3/M
Requesty$3/M
Vertex (Anthropic)$3/M
AIHubMix$3/M
OrcaRouter$3/M
Vertex$3/M
GMI Cloud$3/M
Amazon Bedrock$3/M
Cortecs$3.196/M
Venice AI$3.6/M
Cost calculator
Estimate a workload
$0.00
Excludes taxes, non-token charges, and tiered discounts.
Context limits also differ by provider, from 200K to 1M tokens. Compare the provider table above before choosing on price alone.
Specification
Capabilities
Recorded from the source catalog and provider listings.
Every result stays attached to its source, version, metric and harness. Scores from different versions are never merged, and the profile below plots each benchmark against its own population rather than on a shared scale.
Each axis is this model's percentile among the 7 benchmarks it has published results for, measured against every other model with a score on that same benchmark. Percentiles are used because benchmarks do not share a scale — a 60 on one is not a 60 on another. Hover any point for the raw score.
Humanity's Last Exam
accuracy · 2026-06-30
Source ↗
Terminal-Bench — 2.1
pass@1 · Claude Code
Source ↗
Terminal-Bench#6 of 6
63.1
63.1median 64.971.4
SWE-Atlas Refactoring
score · Claude Code
Source ↗
SWE-Atlas Refactoring#3 of 3
32.21
32.2median 35.648.6
SWE-Atlas Codebase QnA
score · Claude Code
Source ↗
31.22 models scored — too few for a distribution
SWE-Atlas Test Writing
score · Claude Code
Source ↗
31.762 models scored — too few for a distribution
Each strip shows every published score for that benchmark, with this model marked. The lighter shape behind the ticks is the density of results, and the dashed line is the median.
Cost against capability
Price and performance
Listed input price plotted against Humanity's Last Exam, the benchmark with the widest published coverage that this model appears in.
The stepped line is the efficient frontier: at each price, the best score available for that money or less. A model sitting on it is not being beaten by anything cheaper. Price is log-scaled because listed rates span four orders of magnitude. Only models with both a listed price and a score on this benchmark can appear.
Catalog activity
Change log
Field-level changes detected between successful source imports.
Snowflake Cortex
AnyAPI
Kenari
Xpersona
Auriko
Ofox
Merge Gateway
Pioneer
Requesty
Azure Cognitive Services
Abacus
Cloudflare AI Gateway
CrossModel
DigitalOcean
FreeModel
Neon
LLM Gateway
FrogBot
Impossibl
GitHub Copilot
Modelis
Anthropic
OpenRouter
Perplexity Agent
Databricks
FastRouter
DaoXE
AIHubMix
OrcaRouter
GMI Cloud
Amazon Bedrock
Cortecs