Updated daily at 05:00 UTC
Vision-language
Gemini 3.8 Flash
- Released
- 2 Sep 2026
- Status
- Current
- Context
- 1M
- Max output
- 65.5K
- Input per 1M tokens
- $0.75
- Output per 1M tokens
- $3.75
- Open weights
- No
- License
- proprietary
- Parameters
- —
- Knowledge cutoff
- Mar 2026
- Reasoning
- Yes
- Modalities
- text, image, video, audio, pdf → text
Category scores
Out of 100, from independent and third-party results only, by method category-2026-09-29.1.
| Category | Score | Rank | Coverage |
|---|---|---|---|
| Knowledge & reasoning | 60.6 | 7 | 2 of 4 |
| Coding & software engineering | 57.5 | Not ranked | 2 of 5 |
| Vision & multimodal understanding | 2.4 | Not ranked | 1 of 3 |
| Factuality & hallucination | 69.7 | 6 | 1 of 1 |
| Human preference | 57.0 | 11 | 3 of 5 |
Benchmark results
Every result we hold for this model, with who ran it and where it was published.
Knowledge & reasoning
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| Chess Puzzles | Puzzles solved | 61.0% | Independent | high | 2 Sep 2026 | Epoch AI |
| DTBench (Decision Theory Benchmark) | Accuracy (capability questions) | 95.7% | Third party | unknown | — | https://conceptualreasoning.ai/dtbench via Epoch AI |
| DTBench (Decision Theory Benchmark) | Accuracy (capability questions) | 94.9% | Third party | low | — | https://conceptualreasoning.ai/dtbench via Epoch AI |
| DTBench (Decision Theory Benchmark) | Accuracy (capability questions) | 94.9% | Third party | medium | — | https://conceptualreasoning.ai/dtbench via Epoch AI |
| LMCA (Language Model Conceptual Argumentation) | Correlation with expert ratings (Pearson r) | 0.529 | Third party | medium | — | https://conceptualreasoning.ai/lmca via Epoch AI |
| LMCA (Language Model Conceptual Argumentation) | Correlation with expert ratings (Pearson r) | 0.5155 | Third party | unknown | — | https://conceptualreasoning.ai/lmca via Epoch AI |
| LMCA (Language Model Conceptual Argumentation) | Correlation with expert ratings (Pearson r) | 0.5116 | Third party | low | — | https://conceptualreasoning.ai/lmca via Epoch AI |
| Mystery Game Puzzles | Puzzles solved | 47.0% | Independent | high | 2 Sep 2026 | Epoch AI |
Scientific reasoning
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| GPQA Diamond | Accuracy | 95.4% | Independent | high | 2 Sep 2026 | Epoch AI |
| Surface Evolver Bench | Mean score (partial credit) | 76.9% | Third party | high | — | https://yhenon.github.io/surface-evolver-llm-eval/ via Epoch AI |
Mathematics
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| FrontierMath Tier 4 | Accuracy | 22.0% | Independent | high | 2 Sep 2026 | Epoch AI |
| FrontierMath Tiers 1-3 | Accuracy | 68.4% | Independent | high | 2 Sep 2026 | Epoch AI |
| OTIS Mock AIME 2024-2025 | Accuracy | 98.9% | Independent | high | 2 Sep 2026 | Epoch AI |
Coding & software engineering
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| DeepSWE | Pass@1 | 73.8% | Third party | high | — | https://deepswe.datacurve.ai/ via Epoch AI |
| DeepSWE | Pass@1 | 71.0% | Third party | medium | — | https://deepswe.datacurve.ai/ via Epoch AI |
| FrontierCode | Main score (Mean@5) | 41.2% | Third party | medium | — | https://cognition.com/frontiercode via Epoch AI |
Professional knowledge work
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| APEX-Agents | Pass@1 (all rubric criteria met) | 64.3% | Source not stated | unknown | — | Epoch AI via Epoch AI |
Vision & multimodal understanding
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| Furniture Assembly | Score | 31.7% | Independent | high | 10 Sep 2026 | Epoch AI |
Factuality & hallucination
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| SimpleQA Verified | Accuracy (% correct) | 69.7% | Independent | high | 2 Sep 2026 | Epoch AI |
Human preference
| Benchmark | Metric | Value | Provenance | Setting | Evaluated | Source |
|---|---|---|---|---|---|---|
| Code Arena: WebDev (arena.ai) | Arena score (Elo scale) | 1580 | Independent | high | 25 Sep 2026 | Arena |
| Code Arena: WebDev (arena.ai) | Votes | 7338 | Independent | high | 25 Sep 2026 | Arena |
| Text Arena (arena.ai) | Arena score (Elo scale) | 1492 | Independent | high | 25 Sep 2026 | Arena |
| Text Arena (arena.ai) | Votes | 21728 | Independent | high | 25 Sep 2026 | Arena |
| Vision Arena (arena.ai) | Arena score (Elo scale) | 1286 | Independent | high | 28 Sep 2026 | Arena |
| Vision Arena (arena.ai) | Votes | 2711 | Independent | high | 28 Sep 2026 | Arena |
Other Gemini Flash versions
- Gemini 3.8 Flash Cyber16 Sep 2026
- Gemini 3.7 Flash13 Aug 2026
- Gemini 3.6 Flash21 Jul 2026
- Gemini 3.5 Flash19 May 2026
- Gemini 3 Flash17 Dec 2025
- Gemini 2.5 Flash17 Apr 2025
- Gemini 2.0 Flash Thinking19 Dec 2024
- Gemini 2.0 Flash11 Dec 2024
- Gemini 1.5 Flash-8B27 Aug 2024
- Gemini 1.5 Flash14 May 2024
Sources
- https://ai.google.dev/gemini-api/docs/models/gemini-3.8-flash
- https://deepmind.google/models/model-cards/gemini-3-8-flash/
- https://ai.google.dev/gemini-api/docs/latest-model
- https://ai.google.dev/gemini-api/docs/pricing
- https://ai.google.dev/gemini-api/docs/deprecations
- https://ai.google.dev/gemini-api/docs/changelog