Model comparison
Claude Haiku 4.5 vs Gemini 3.1 Flash Lite
Gemini 3.1 Flash Lite is the stronger model overall, scoring 40.8 to 39.5 on the Noometry Index.
Last verified . 32 shared benchmarks.
GoogleSummary
- They share 32 benchmarks with published results for both. Claude Haiku 4.5 scores higher in 4 categories and Gemini 3.1 Flash Lite in 6 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in multimodal, where Gemini 3.1 Flash Lite leads 39.4 to 26.8.
- The biggest single-benchmark swing is Chess Puzzles: 8% for Claude Haiku 4.5 and 25% for Gemini 3.1 Flash Lite.
- Gemini 3.1 Flash Lite is cheaper at $0.25 / $1.50 per million input/output tokens, against $1 / $5 for Claude Haiku 4.5.
- Gemini 3.1 Flash Lite accepts more context: 1.05M tokens versus 200K.
Side by side
| Claude Haiku 4.5 | Gemini 3.1 Flash Lite | |
|---|---|---|
| Provider | Anthropic | |
| Noometry Index | 39.5 | 40.8 |
| Released | 2025-10-15 | 2026-03-03 |
| Weights | Proprietary | Proprietary |
| Context window | 200K | 1.05M |
| Max output | 64K | 66K |
| Input $ / M tokens | $1 | $0.25 |
| Output $ / M tokens | $5 | $1.50 |
| Results tracked | 53 | 38 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Haiku 4.5 leads
Claude Haiku 4.5: 44.0 (#78), Gemini 3.1 Flash Lite: 37.8 (#188)
| Benchmark | Claude Haiku 4.5 | Gemini 3.1 Flash Lite |
|---|---|---|
| LMArena WebDev | 1330 | 1256 |
| SciCode | 43.3% | 41.9% |
| WeirdML | 45.4% | 52.2% |
| LMArena Coding | 1453 | 1400 |
| ALE-Bench | 653.48 | 797.73 |
| SWE-bench Verified (bash only) | 66.6% | — |
| SWE-bench Multilingual | 64.7% | — |
Agentic & Tool Use Claude Haiku 4.5 leads
Claude Haiku 4.5: 33.6 (#52), Gemini 3.1 Flash Lite: 30.2 (#79)
| Benchmark | Claude Haiku 4.5 | Gemini 3.1 Flash Lite |
|---|---|---|
| DeepResearch Bench | 45.5% | 37.3% |
| Terminal-Bench | 35.5% | — |
| Berkeley Function Calling Leaderboard | 68.7% | — |
| BALROG | 31.2% | — |
| ExploitBench | 13.7% | — |
| Vending-Bench 2 | 458.89 | — |
Reasoning Gemini 3.1 Flash Lite leads
Claude Haiku 4.5: 15.1 (#320), Gemini 3.1 Flash Lite: 22.9 (#186)
| Benchmark | Claude Haiku 4.5 | Gemini 3.1 Flash Lite |
|---|---|---|
| NYT Connections (extended) | 14.3% | 8.2% |
| CritPt | 0% | 1.1% |
| Chess Puzzles | 8% | 25% |
| LMArena Hard Prompts | 1420 | 1407 |
| DTBench | 73.6% | 76.8% |
| LMCA | 30.9% | 35% |
| Epoch Capabilities Index | 142.41 | 144.47 |
| ForecastBench | 61.4 | 54.4 |
| ARC-AGI-2 | 4% | — |
| Kagi LLM Benchmark | — | 67.2% |
| ARC-AGI-1 | 47.7% | — |
| EnigmaEval | — | 3% |
| Thematic Generalization | — | 63.3% |
Math Claude Haiku 4.5 leads
Claude Haiku 4.5: 44.9 (#78), Gemini 3.1 Flash Lite: 40.7 (#90)
| Benchmark | Claude Haiku 4.5 | Gemini 3.1 Flash Lite |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 66.7% | 80% |
| LMArena Math | 1396 | 1428 |
| FrontierMath (Tiers 1-3) | — | 27.7% |
| Omni-MATH | 56.1% | — |
| MATH Level 5 | 96.4% | — |
| FrontierMath (Feb 2025 set) | 5.9% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge Gemini 3.1 Flash Lite leads
Claude Haiku 4.5: 37.7 (#153), Gemini 3.1 Flash Lite: 41.9 (#104)
| Benchmark | Claude Haiku 4.5 | Gemini 3.1 Flash Lite |
|---|---|---|
| GPQA Diamond | 71.2% | 81.8% |
| Vectara Hallucination Rate | 9.8% | 8.2% |
| LMArena Expert | 1442 | 1398 |
| Humanity's Last Exam | — | 8.6% |
| SimpleQA Verified | 13.2% | — |
| MMLU-Pro | 77.7% | — |
| GPQA (HELM) | 60.5% | — |
Multimodal Gemini 3.1 Flash Lite leads
Claude Haiku 4.5: 26.8 (#118), Gemini 3.1 Flash Lite: 39.4 (#60)
| Benchmark | Claude Haiku 4.5 | Gemini 3.1 Flash Lite |
|---|---|---|
| LMArena Vision | — | 1240 |
| Blueprint-Bench 2 | 0% | — |
| LMArena Document | 1420 | — |
Multilingual Gemini 3.1 Flash Lite leads
Claude Haiku 4.5: 49.9 (#129), Gemini 3.1 Flash Lite: 52.3 (#86)
| Benchmark | Claude Haiku 4.5 | Gemini 3.1 Flash Lite |
|---|---|---|
| LMArena Non-English | 1377 | 1411 |
| LMArena Chinese | 1417 | 1461 |
| LMArena French | 1408 | 1424 |
| LMArena German | 1375 | 1429 |
| LMArena Japanese | 1339 | 1413 |
| LMArena Korean | 1347 | 1392 |
| LMArena Russian | 1381 | 1420 |
| LMArena Spanish | 1420 | 1421 |
Instruction Following Gemini 3.1 Flash Lite leads
Claude Haiku 4.5: 71.4 (#149), Gemini 3.1 Flash Lite: 72.7 (#131)
| Benchmark | Claude Haiku 4.5 | Gemini 3.1 Flash Lite |
|---|---|---|
| LMArena Instruction Following | 1414 | 1377 |
| IFEval | 80.1% | — |
Long Context Claude Haiku 4.5 leads
Claude Haiku 4.5: 43.6 (#92), Gemini 3.1 Flash Lite: 42.5 (#122)
| Benchmark | Claude Haiku 4.5 | Gemini 3.1 Flash Lite |
|---|---|---|
| LMArena Longer Query | 1427 | 1394 |
Writing & Preference Gemini 3.1 Flash Lite leads
Claude Haiku 4.5: 57.9 (#123), Gemini 3.1 Flash Lite: 60.9 (#94)
| Benchmark | Claude Haiku 4.5 | Gemini 3.1 Flash Lite |
|---|---|---|
| LMArena Text | 1396 | 1416 |
| LMArena Creative Writing | 1372 | 1401 |
| LMArena Multi-Turn | 1409 | 1417 |
| WildBench | 83.9% | — |
| EQ-Bench 4 | 1064 | — |
Frequently asked questions
Is Claude Haiku 4.5 better than Gemini 3.1 Flash Lite?
Gemini 3.1 Flash Lite is the stronger model overall, scoring 40.8 to 39.5 on the Noometry Index.
Which is cheaper, Claude Haiku 4.5 or Gemini 3.1 Flash Lite?
Gemini 3.1 Flash Lite is cheaper. It lists at $0.25 per million input tokens and $1.50 per million output tokens; Claude Haiku 4.5 lists at $1 and $5.
Is Claude Haiku 4.5 or Gemini 3.1 Flash Lite better for coding?
Claude Haiku 4.5 scores higher on coding benchmarks: 44.0 versus 37.8 in the Noometry coding category.
Which has the bigger context window?
Gemini 3.1 Flash Lite does, with 1.05M tokens against 200K.
How many benchmarks do Claude Haiku 4.5 and Gemini 3.1 Flash Lite share?
32 benchmarks have published results for both models. Claude Haiku 4.5 has 53 scored results on Noometry and Gemini 3.1 Flash Lite has 38.