Model comparison
Claude 3 Haiku vs o1
Head-to-head evidence from 11 shared benchmark results across 5 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
Public leaderboard positions: Claude 3 Haiku #204 (Supported); o1 #136 (Estimated). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. Claude 3 Haiku and o1 share 11 comparable benchmark results. 0 of 8 categories are comparable. 1 result is unique to Claude 3 Haiku; 5 to o1.
Updated July 24, 2026- Shared results
- 11
- Claude 3 Haiku only
- 1
- o1 only
- 5
- Comparable categories
- 0 / 8
Benchmark data for Claude 3 Haiku and o1 is coming soon on BenchLM.
Confidence note. This is a partial-evidence comparison with 11 shared benchmark results across 5 evidence categories; 0 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.
Why this result
BenchLM has partial data for these models, but not enough overlapping benchmark coverage to produce a fair score-level comparison yet.
o1 is priced at $15.00 input / $60.00 output per 1M tokens, versus $0.25 input / $1.25 output per 1M tokens for Claude 3 Haiku.
Category breakdown
Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.
| Category | Claude 3 Haiku | Δ | o1 |
|---|---|---|---|
| Knowledge | Claude 3 HaikuNot measured | MarginNo overlap | o175.7 |
| Math | Claude 3 HaikuNot measured | MarginNo overlap | o19.3 |
| Inst. Following | Claude 3 HaikuNot measured | MarginNo overlap | o192.2 |
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | Claude 3 Haiku | o1 | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | Claude 3 Haiku$0.25 input / $1.25 output | o1$15 input / $60 output | Claude 3 Haiku has the lower combined listed price. |
| Generation speedtokens per second | Claude 3 Haiku138 tok/s | o198 tok/s | Claude 3 Haiku has the higher measured throughput. |
| First-answer latencyseconds to first token | Claude 3 Haiku1.16 s | o132.29 s | Claude 3 Haiku reaches the first token sooner. |
| Context windowmaximum listed tokens | Claude 3 Haiku200K | o1200K | Listed context windows are equal. |
Benchmark Deep Dive
Agentic1 benchmarks
| Benchmark | Claude 3 Haiku | o1 | Result |
|---|---|---|---|
| τ²-bench resultsSource | 21.1% | 62.6% | o1 leads |
Coding2 benchmarks
Reasoning2 benchmarks
Knowledge8 benchmarks
| Benchmark | Claude 3 Haiku | o1 | Result |
|---|---|---|---|
| Artificial Analysis Intelligence IndexSource | 3.9% | 23.4% | o1 leads |
| AA-GPQA DiamondSource | 37.4% | 74.7% | o1 leads |
| AA-HLESource | 3.9% | 7.7% | o1 leads |
| AA-Omniscience IndexSource | -47.6% | -10.5% | o1 leads |
| AA-Omniscience AccuracySource | 17.2% | 34.7% | o1 leads |
| AA-Omniscience Hallucination RateSource | 78.2% | 69.3% | o1 leads |
| MMLUSource | — | 91.8% | Not comparable |
| GPQASource | — | 75.7% | Not comparable |
Math1 benchmarks
| Benchmark | Claude 3 Haiku | o1 | Result |
|---|---|---|---|
| FrontierMath v2 (Tiers 1-3)Source | — | 9.310% | Not comparable |
Multimodal1 benchmarks
| Benchmark | Claude 3 Haiku | o1 | Result |
|---|---|---|---|
| AA-MMMU-ProSource | 30.8% | — | Not comparable |
Frequently Asked Questions (3)
Can I compare Claude 3 Haiku and o1 on BenchLM yet?
Not fully yet. BenchLM is tracking both models, but the sourced benchmark breakdown for this comparison is still coming soon.
Why does this comparison show “coming soon”?
BenchLM only shows category winners and benchmark-level calls when we have sourced results that can be compared fairly. For these models, the public benchmark coverage is not complete enough yet.
What data is available for Claude 3 Haiku and o1 today?
Claude 3 Haiku: $0.25 input / $1.25 output per 1M tokens o1: $15.00 input / $60.00 output per 1M tokens Both model pages still include creator, context window, reasoning mode, and other metadata while benchmark coverage fills in.