LLM API Pricing & Performance (July 2026)

Cross-vendor comparison of standard API pricing (input/output per 1M tokens) and Artificial Analysis capability scores for major LLMs. All figures sourced, retrieved 22 July 2026.

ProviderModelInput (per 1M)Output (per 1M)Score
OpenAIGPT-5.6 Sol$5.00$30.0059
AnthropicClaude Opus 4.8$5.00$25.0056
xAIGrok 4.5$2.00$6.0054
AnthropicClaude Sonnet 5$2.00*$10.00*53
OpenAIGPT-5.6 Luna$1.00$6.0051
GoogleGemini 3.6 Flash$1.50$7.5050
GoogleGemini 3.1 Pro$2.00$12.0046
DeepSeekDeepSeek V4 Pro$0.44$0.8744
DeepSeekDeepSeek V4 Flash$0.14$0.2840
AnthropicClaude Haiku 4.5$1.00$5.0024

Method & sources

Pricing is from each provider''s official pricing page — standard (non-batch, non-cached) per-1M-token rates in USD. Capability scores are the Artificial Analysis Intelligence Index (July 2026 snapshot); reasoning-effort settings can shift scores. LLM pricing changes frequently and often carries promotional periods and long-context tiers; bill against the official pages.

Source: https://platform.claude.com/docs/en/docs/about-claude/pricing

Retrieved: 2026-07-22

→ Read the full analysis