Google's most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows
Gemini 3.8 Flash by google is offered by 37 providers on this page. Public prices are shown for 36 of them. Its official list price is $0.75 per 1M input tokens and $3.75 per 1M output tokens. No paid channel is priced below the official list price. Prices vary about 2.0× across the providers that host it.
Artificial Analysis rates it 40.9 on the Intelligence Index, with 76.3 for coding and 40.2 for agentic tasks. Against its lowest blended price of $1.50 per 1M, that is roughly 27 index points per dollar, which is the value ratio the leaderboards rank on. Median output speed is 147 tokens per second, with 22.34s to the first token. Running one task of the Artificial Analysis suite costs about $1.2428, which reflects how many tokens its reasoning consumes rather than the unit price alone.
The context window is 1,048,576 tokens at the reference host, but hosts report different limits, from 1,000,000 to 1,048,576, so the usable window depends on the provider you pick. It supports reasoning, tool use, and structured output. Accepted input modalities are Text, Image, Video, Audio, and PDF.
Quality is independently evaluated by Artificial Analysis. Speed/latency are model-level medians.
| Provider | Tier | Input | Output | Cache read | Cache write | Context | Output limit | Status |
|---|---|---|---|---|---|---|---|---|
| GoogleOfficial | First-party | $0.75 | $3.75 | $0.075 | — | 1,048,576 | 65,536 | |
| VertexOfficial | Cloud | $0.75 | $3.75 | $0.075 | — | 1,048,576 | 65,536 | |
| NanoGPT | Gateway | $0.75 | $3.75 | $0.075 | $0.042 | 1,048,576 | 65,536 | |
| NEAR AI Cloud | Gateway | $0.75 | $3.75 | $0.075 | — | 1,048,576 | 65,536 | |
| LLM Gateway | Gateway | $0.75 | $3.75 | $0.075 | $0.083 | 1,048,576 | 65,536 | |
| Kilo Gateway | Gateway | $0.75 | $3.75 | $0.075 | $0.042 | 1,048,576 | 65,536 | |
| 302.AI | Gateway | $0.75 | $3.75 | — | — | 1,048,576 | 65,536 | |
| Vercel AI Gateway | Cloud | $0.75 | $3.75 | $0.075 | — | 1,000,000 ⚠ | 65,535 | |
| Ofox | Gateway | $0.75 | $3.75 | $0.075 | $0.042 | 1,000,000 ⚠ | 65,536 | |
| AIHubMix | Gateway | $0.75 | $3.75 | $0.075 | — | 1,048,576 | 65,536 | |
Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.
Interleaved thinking (reasoning between tool calls) is declared by 1 of 37 providers.
4 / 37 providers expose no reasoning control (reasoning_options: []).
| Provider | Audio in | Audio out |
|---|---|---|
| $0.751.0× | — | |
| Vertex | $0.751.0× | — |
| Ofox | $0.751.0× | — |
| Tempr Gateway | $0.751.0× | — |
| Eden AI | $1.501.0× | — |
| Eden AI | $1.501.0× | — |
| Eden AI | $1.501.0× | — |
| Eden AI | $1.501.0× | — |
| Eden AI | $1.501.0× | — |
| Eden AI | $1.501.0× | — |
| OpenCode | $1.501.0× | — |
Audio tokens are priced on a separate scale from text tokens, so these are not comparable with the text prices above. The multiplier compares each price with the same provider's text price for the same direction.
Price history accumulates from each data sync; currently only 1 sample(s) (2026-09-03). Each future sync adds a point, and once accumulated a line is drawn here.
| Name | Conditions | Score | Metric | Source |
|---|---|---|---|---|
| DeepSWE | harness: mini-swe-agent · variant: high thinking · v1.1 | 73.7 | pass@1 | Source ↗ |
| GDPval-AA | v2 | 1545 | Elo | Source ↗ |
| Finance Agent | v2 | 61.4 | pass@1 | Source ↗ |
| Harvey Legal Agent Benchmark | — | 10 | all-pass rate | Source ↗ |
| Terminal-Bench | harness: Terminus 2 · v2.1 | 89.4 | pass@1 | Source ↗ |
| Terminal-Bench | v4.0 | 19.1 | pass@1 | Source ↗ |
| GDP.PDF | — | 35 | all-pass rate | Source ↗ |
| CharXiv Reasoning | variant: without tools | 86.2 | pass@1 | Source ↗ |
| LVBench | variant: agentic | 87.8 | pass@1 | Source ↗ |
| LVBench | variant: static; 1024 frames; without tools | 87.1 | pass@1 | Source ↗ |
| HLE-Verified | dataset: 1811 verified and revised items | 54.9 | accuracy | Source ↗ |
| OSWorld | harness: Gemini CUA · variant: batched tool calls; best of 3 runs; 500 steps · dataset: before 2026-08-08 patch · v2.0 | 59 | partial score | Source ↗ |
| BioMysteryBench | variant: with tools · dataset: human-solvable | 88.8 | pass@1 | Source ↗ |
| BioMysteryBench | variant: with tools · dataset: human-difficult | 56.5 | pass@1 | Source ↗ |
| LAB-Bench | variant: with tools · dataset: 11 subtasks · v2 | 86.2 | macro-average | Source ↗ |
The same benchmark scores very differently across harness / dataset, so the qualifying conditions must be shown together.