Compare / head-to-head
Claude Haiku 4.5vsGemini 3.1 Flash-Lite
Gemini 3.1 Flash-Lite leads 3 of 3 shared benchmarks. Gemini 3.1 Flash-Lite is 3.4x cheaper per token. Gemini 3.1 Flash-Lite has the larger context window (1.0M tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
0 – 3
Gemini 3.1 Flash-Lite leads
Cheaper per token
Gemini 3.1 Flash-Lite
3.4x cheaper, input + output
Larger context
Gemini 3.1 Flash-Lite
1.0M tokens
Provider uptime (30d)
100% · 100%
Anthropic · Google
Claude Haiku 4.5
Anthropic · released 2025-10-15
textvision
Gemini 3.1 Flash-Lite
Google · released 2026-03-03
textvisionaudio
Quality
Benchmark matrix
3 shared · 1 only Claude Haiku 4.5 · 6 only Gemini 3.1 Flash-Lite| Benchmark | Claude Haiku 4.5 | Gemini 3.1 Flash-Lite | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 3 | ||||
| GPQA Diamond | 71.2% ↗epoch | 86.9% ↗ | -15.7 pt | Gemini 3.1 Flash-Lite |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 66.7% ↗epoch | 80% ↗epoch | -13.3 pt | Gemini 3.1 Flash-Lite |
| SimpleQA Verified | 13.2% ↗epoch | 43.3% ↗ | -30.1 pt | Gemini 3.1 Flash-Lite |
| Only Claude Haiku 4.5 reports · 1 | ||||
| SWE-bench Verified | 73.3% ↗ | not reported | — | — |
| Only Gemini 3.1 Flash-Lite reports · 6 | ||||
| Humanity's Last Exam (no tools) | not reported | 16% ↗ | — | — |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | not reported | 27.7% ↗epoch | — | — |
| LiveCodeBench | not reported | 72% ↗ | — | — |
| MMMLU | not reported | 88.9% ↗ | — | — |
| MMMU-Pro | not reported | 76.8% ↗ | — | — |
| MRCR v2 (128k average) | not reported | 60.1% ↗ | — | — |
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is Claude Haiku 4.5 minus Gemini 3.1 Flash-Lite in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | Claude Haiku 4.5 | Gemini 3.1 Flash-Lite | Edge |
|---|---|---|---|
| Context window | 200K tokens | 1.0M tokens | Gemini 3.1 Flash-Lite |
| Max output | 64K tokens | 66K tokens | Gemini 3.1 Flash-Lite |
| Input price / 1M | $1 ↗ | $0.25 ↗ | Gemini 3.1 Flash-Lite |
| Output price / 1M | $5 ↗ | $1.5 ↗ | Gemini 3.1 Flash-Lite |
| Cached input / 1M | $0.1 ↗ | $0.025 ↗ | Gemini 3.1 Flash-Lite |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $6 | $1.75 | Gemini 3.1 Flash-Lite |
| Modalities | text · vision | text · vision · audio | Gemini 3.1 Flash-Lite |
| Released | 2025-10-15 | 2026-03-03 | — |
| Cited benchmark scores | 4 | 10 | — |
Reliability
Provider status
Last 60 days · refreshed daily on this page · live on /statusMore matchups
Claude Haiku 4.5 vs …
Models sharing the most benchmarksMore matchups
Gemini 3.1 Flash-Lite vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Claude Haiku 4.5 and Gemini 3.1 Flash-Lite on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.