Compare / head-to-head
Claude Opus 4.8vs
Gemini 3 Flash Preview
Claude Opus 4.8 leads 15 of 18 shared benchmarks. Gemini 3 Flash Preview has the larger context window (1.0M tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
15 – 3
Claude Opus 4.8 leads
Cheaper per token
—
list price, input + output
Larger context
Gemini 3 Flash Preview
1.0M tokens
Providers
2 providers
Anthropic · Google
Claude Opus 4.8
Anthropic · released 2026-05-28
textvision
Gemini 3 Flash Preview
Google · released 2025-12-17
textvisionaudio
- Context
- 1.0M
- Max out
- 66K
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 34 · 12 core
Quality
Benchmark matrix
18 shared · 22 only Claude Opus 4.8 · 8 only Gemini 3 Flash Preview| Benchmark | Claude Opus 4.8 | Gemini 3 Flash Preview | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 18 | ||||
| SWE-bench Verified | 88.6% ↗ | 78% ↗ | +10.6 pt | Claude Opus 4.8 |
| GPQA Diamond | 93.6% ↗ | 90.4% ↗ | +3.2 pt | Claude Opus 4.8 |
| Humanity's Last Exam (no tools) | 49.8% ↗ | 33.7% ↗ | +16.1 pt | Claude Opus 4.8 |
| LMArena Elo | 1452.7 ↗ | 1472.7 ↗ | -20 | Gemini 3 Flash Preview |
| Chess Puzzles (Epoch AI run) | 34% ↗epoch run | 40% ↗epoch run | -6 pt | Gemini 3 Flash Preview |
| FrontierMath Tier 4 v2 (Epoch AI run) | 56.1% ↗epoch run | 17.1% ↗epoch run | +39 pt | Claude Opus 4.8 |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | 80% ↗epoch run | 51.2% ↗epoch run | +28.8 pt | Claude Opus 4.8 |
| GSO Opt@1 (GSO) | 47.1% ↗ | 7.8% ↗ | +39.3 pt | Claude Opus 4.8 |
| LMArena Vision (LMArena) | 1286.4 ↗ | 1272.9 ↗ | +13.5 | Claude Opus 4.8 |
| LMArena WebDev (LMArena) | 1555.5 ↗ | 1438.9 ↗ | +116.6 | Claude Opus 4.8 |
| Mystery Game Puzzles (Epoch AI run) | 36% ↗epoch run | 26% ↗epoch run | +10 pt | Claude Opus 4.8 |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 98.3% ↗epoch run | 95.6% ↗epoch run | +2.7 pt | Claude Opus 4.8 |
| SAGE (Vals AI) | 54.8% ↗ | 51.8% ↗ | +3 pt | Claude Opus 4.8 |
| SimpleBench (SimpleBench) | 64.8% ↗ | 61.1% ↗ | +3.7 pt | Claude Opus 4.8 |
| SimpleQA Verified | 53% ↗epoch run | 68.7% ↗ | -15.7 pt | Gemini 3 Flash Preview |
| tau2-bench Banking Knowledge (Sierra) | 39.7% ↗ | 27.3% ↗ | +12.4 pt | Claude Opus 4.8 |
| Vending-Bench 2 (Andon Labs) | 5787.43 ↗ | 3634.72 ↗ | +2152.7 | Claude Opus 4.8 |
| WeirdML (Håvard Tveit Ihle) | 82.9% ↗ | 61.6% ↗ | +21.3 pt | Claude Opus 4.8 |
| Only Claude Opus 4.8 reports · 22 | ||||
| AIME 2026 | 100% ↗matharena ⚠ | not reported | — | — |
| APEX-Agents (Mercor) | 48.9% ↗ | not reported | — | — |
| BrowseComp | 84.3% ↗ | not reported | — | — |
| DeepSWE v1.1 (Datacurve) | 59% ↗ | not reported | — | — |
| EBR-bench (Epoch AI run) | 28.6% ↗epoch run | not reported | — | — |
| Furniture Assembly (Epoch AI run) | 42.5% ↗epoch run | not reported | — | — |
| Humanity's Last Exam (with tools) | 57.9% ↗ | not reported | — | — |
| LiveBench Agentic Coding (LiveBench) | 50.5% ↗ | not reported | — | — |
| LiveBench Coding (LiveBench) | 81.8% ↗ | not reported | — | — |
| LiveBench Data Analysis (LiveBench) | 66% ↗ | not reported | — | — |
| LiveBench Instruction Following (LiveBench) | 72% ↗ | not reported | — | — |
| LiveBench Language (LiveBench) | 79.7% ↗ | not reported | — | — |
| LiveBench Mathematics (LiveBench) | 94.3% ↗ | not reported | — | — |
| LiveBench Reasoning (LiveBench) | 89.2% ↗ | not reported | — | — |
| LMArena Agent (LMArena) | 0.0664 ↗ | not reported | — | — |
| OSWorld-Verified | 83.4% ↗ | not reported | — | — |
| SWE-bench Multilingual | 84.4% ↗ | not reported | — | — |
| SWE-bench Multimodal | 38.4% ↗ | not reported | — | — |
| SWE-Bench Pro | 69.2% ↗ | not reported | — | — |
| Terminal-Bench 2.1 | 74.6% ↗ | not reported | — | — |
| Terminal-Bench 4.0 (Vals AI) | 23.2% ↗ | not reported | — | — |
| Toolathlon-Verified (HKUST) | 76.2% ↗ | not reported | — | — |
| Only Gemini 3 Flash Preview reports · 8 | ||||
| AIME 2025 | not reported | 95.2% ↗ | — | — |
| BALROG (BALROG) | not reported | 48.1% ↗ | — | — |
| MMMU-Pro | not reported | 81.2% ↗ | — | — |
| SWE-bench Verified (Epoch AI run) | not reported | 75.4% ↗epoch run | — | — |
| tau2-bench Airline (Sierra) | not reported | 82.5% ↗ | — | — |
| tau2-bench Retail (Sierra) | not reported | 76.8% ↗ | — | — |
| tau2-bench Telecom (Sierra) | not reported | 91.2% ↗ | — | — |
| Terminal-Bench 2.0 | not reported | 47.6% ↗ | — | — |
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 Opus 4.8 minus Gemini 3 Flash Preview 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 Opus 4.8 | Gemini 3 Flash Preview | Edge |
|---|---|---|---|
| Context window | 1M tokens | 1.0M tokens | Gemini 3 Flash Preview |
| Max output | 128K tokens | 66K tokens | Claude Opus 4.8 |
| Input price / 1M | $5 ↗ | — | — |
| Output price / 1M | $25 ↗ | — | — |
| Cached input / 1M | $0.5 ↗ | — | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $30 | — | — |
| Modalities | text · vision | text · vision · audio | Gemini 3 Flash Preview |
| Released | 2026-05-28 | 2025-12-17 | — |
| Cited benchmark scores | 47 | 34 | — |
Reliability
Provider status
Live from /statusMore matchups
Claude Opus 4.8 vs …
Models sharing the most benchmarksMore matchups
Gemini 3 Flash Preview vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Claude Opus 4.8 and Gemini 3 Flash Preview on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.