Output tokens cost $25.00 per million on anthropic-opus-4-6 and $0.28 per million on claude-3-haiku-20240229. Input tokens cost $5.00 per million on anthropic-opus-4-6 and $0.28 per million on claude-3-haiku-20240229. Cached input tokens are billed at $0.50 per million on anthropic-opus-4-6 and $0.28 per million on claude-3-haiku-20240229. Time to first token (TTFT) measured on AIHubMix is 3.1s on anthropic-opus-4-6 and 0.8s on claude-3-haiku-20240229. Measured output throughput is 39.7 tok/s on anthropic-opus-4-6 and 67.2 tok/s on claude-3-haiku-20240229. On the LMArena coding leaderboard anthropic-opus-4-6 scores 1548 and claude-3-haiku-20240229 scores 1300.
anthropic-opus-4-6claude-3-haiku-20240229Claude Opus 4.6 is Anthropic’s latest state-of-the-art reasoning model. It features an adaptive “thinking” mode that dynamically decides when to think and how much to think. At the default effort level (high), Claude will almost always engage in thinking. At lower effort levels, it may skip thinking for simple problems. ⚠️ The minimum cache token for claude-opus-4-6 has been increased from 1,024 to 4,096 tokens.
Prices are per million tokens. Time to First Token and throughput are rolling averages measured on AIHubMix.
Promotional prices show the discounted rate; see each model page for promotion windows.
Daily traffic served through AIHubMix — how demand for each model is trending.
Measured on real AIHubMix traffic, hourly buckets. Gaps mean no traffic in that hour.
LMArena ratings by capability (Bradley-Terry, commonly called Elo). Higher is better.
Source: LMArena (arena.ai) leaderboard, imported by AIHubMix. Models without published ratings are omitted per chart.
Estimate your monthly bill for the same workload on each model.
Monthly = daily × 30. Discounted rates applied where a promotion is active.
claude-3-haiku-20240229: $0.28/M output tokens; anthropic-opus-4-6: $25.00/M. Use the cost calculator above to estimate your own workload.
anthropic-opus-4-6: 1548; claude-3-haiku-20240229: 1300 (LMArena coding leaderboard).
claude-3-haiku-20240229: 0.8s time to first token measured on AIHubMix; see the live performance charts above for how each model behaves across the day.
claude-3-haiku-20240229 at 67.2 tok/s and anthropic-opus-4-6 at 39.7 tok/s, measured as output throughput on AIHubMix — a separate metric from time to first token.
anthropic-opus-4-6 accepts text and image input and supports thinking, tool calling, function calling and structured outputs; claude-3-haiku-20240229 accepts text and image input.
Yes. AIHubMix serves every model on this page behind one OpenAI-compatible endpoint, so switching between them is a one-line change to the model field — no second account, key or SDK.
Related model match-ups readers also look at.
AIHubMix