Gemini 3.8 Flash
Googlefamily · Gemini 3.8ai.google.dev/gemini-api/docs/models/gemini-3.8-
Our most intelligent Flash model, engineered for long-horizon software engineering, autonomous agents, and complex enterprise workflows.
Updated 24 h ago · first seen 11 Sept 2026
model_01M293V7SGMT3E44NWTHC33FJ1
Overview
Identity
- Canonical model
- Yesidentity confidence: highOne row per real model release. Artifacts (checkpoints, quantisations, conversions) and folded evaluation variants point here.
- Official checkpoints
- None recorded — closed weightsofficial_checkpoints = hf_repo identifiers carried by the model itself; artifacts are separate entities pointing here through canonical_id.
- Artifacts
- None recordedSeparate entities (checkpoint · quantization · conversion · packaging) pointing to this model through canonical_id.
- Provider deployments
- 2
- API aliases
- gemini-3-8-flashgemini-3-8-flash-lowgemini-3-8-flash-mediumgemini-3.8-flashgoogle/gemini-3.8-flashIdentifiers under which providers and evaluators refer to this model.
- Folded evaluation variants
- 2Effort / thinking variants (…-high, …-non-reasoning) are result configurations of this model, not separate models. Their old URLs redirect here.
Openness
Closed / proprietary— weights not available; 7 dimensions unknown.
Weights are not available; the model is reachable only through an API or a product.
Weights
No
Inference code
—
Training code
—
Training data
—
Dataset
—
Commercial use
—
Redistribution
—
Derivatives
—
dimensions marked null are unknown, not false
Key facts
- Release date
Source:OpenRouter public model & pricing listingT2observed 5 d agomedium
- Status
Source:Google AI for Developers (Gemini API docs)T1observed 5 d agohigh
- Version
Source:Google AI for Developers (Gemini API docs)T2observed 24 h agomediumLLM-extracted
- Official page
Source:Google AI for Developers (Gemini API docs)T1observed 5 d agohigh
- API model id
Source:Google AI for Developers (Gemini API docs)T1observed 5 d agohigh
- Openrouter id
Source:OpenRouter public model & pricing listingT2observed 5 d agomedium
Architecture
- Tokenizer
Source:OpenRouter public model & pricing listingT2observed 5 d agomedium
Capabilities
Modalities
- Modalities
- audiodocumentimagetextvideo
- Input
- audiodocumentimagetextvideo
- Output
- text
Capabilities
Tool calling
Yes
Google AI for Developers (Gemini API docs) · T1
Structured output
Yes
Google AI for Developers (Gemini API docs) · T1
Reasoning
Yes
Google AI for Developers (Gemini API docs) · T1
Vision
Yes
Google AI for Developers (Gemini API docs) · T1
Audio
Yes
Google AI for Developers (Gemini API docs) · T1
Fine-tuning available
Unavailable
- Context window
Source:Google AI for Developers (Gemini API docs)T1observed 4 d agohigh
- Max output
Source:Google AI for Developers (Gemini API docs)T1observed 5 d agohigh
- Tokenizer
Source:OpenRouter public model & pricing listingT2observed 5 d agomedium
Benchmarks50
Compare with another model →Comparable same task and conditions · Partially comparable same task, conditions differ (effort, temperature, judge) · Not comparable different variant or metric
Current rows only, grouped by benchmark → canonical metric → comparability group (task configuration). Effort variants folded into this model appear as rows of the same group. 50 current rows in total. “vs leader” compares with the current leader of the benchmark's primary group only; other groups are not directly comparable. Comparability rules →
Providers & Pricing6
All offers in the price terminal →USD per 1M tokens as published by each provider; native units (per-request fees, flex/priority tiers) are kept verbatim. Rows are append-only — every price change is kept in the history below. Cost of a workload →
Price history
Output price · USD / 1M tokens 2 providers
- Google Gemini API
- OpenRouter
- Google Gemini API$1.88 → $3.7515 Sept 2026
- OpenRouter$3.75 → $1.8812 Sept 2026
- OpenRouterfirst observed $3.7512 Sept 2026
- Google Gemini API$3.75 → $1.8811 Sept 2026
- Google Gemini APIfirst observed $3.7511 Sept 2026
Input price · USD / 1M tokens 2 providers
- Google Gemini API
- OpenRouter
- Google Gemini API$0.375 → $0.7515 Sept 2026
- OpenRouter$0.75 → $0.37512 Sept 2026
- OpenRouterfirst observed $0.7512 Sept 2026
- Google Gemini API$0.75 → $0.37511 Sept 2026
- Google Gemini APIfirst observed $0.7511 Sept 2026
Versions & Artifacts0
Version history
Context window2 changes
11 Sept 2026→11 Sept 2026→12 Sept 2026current
Max outputfirst observation only
11 Sept 2026current
Opennessfirst observation only
11 Sept 2026current
Statusfirst observation only
11 Sept 2026current
Each hop is a claim: click a value for its source, tier and observation time. Nothing is overwritten — a new observation closes the previous claim.
Artifacts 0
No artifact (checkpoint, quantisation, conversion or packaging) points to this model yet.
Timeline27
Full timeline →Google Gemini API lists Gemini 3.8 Flash at $0.75 in / $3.75 out per 1M tokens
googleOpenRouter lists Gemini 3.8 Flash at $0.375 in / $1.875 out per 1M tokens
openrouterOpenRouter lists Gemini 3.8 Flash at $0.75 in / $3.75 out per 1M tokens
openrouterGemini 3.8 Flash scores 83.9% on Terminal-Bench
artificial_analysisGemini 3.8 Flash scores 19.7% on Terminal-Bench
artificial_analysisGemini 3.8 Flash scores 84.22% on MMMU-Pro
artificial_analysisGemini 3.8 Flash scores 55.09% on SciCode
artificial_analysisGemini 3.8 Flash scores 42.12% on Humanity's Last Exam
artificial_analysisGemini 3.8 Flash scores 93.54% on GPQA Diamond
artificial_analysisGemini 3.8 Flash scores 40 on Artificial Analysis Intelligence Index
artificial_analysisGemini 3.8 Flash scores 87.64% on Terminal-Bench
artificial_analysisGemini 3.8 Flash scores 19.7% on Terminal-Bench
artificial_analysisGemini 3.8 Flash scores 85.61% on MMMU-Pro
artificial_analysisGemini 3.8 Flash scores 56.6% on SciCode
artificial_analysisGemini 3.8 Flash scores 47.82% on Humanity's Last Exam
artificial_analysisGemini 3.8 Flash scores 95.25% on GPQA Diamond
artificial_analysisGemini 3.8 Flash scores 41.19 on Artificial Analysis Intelligence Index
artificial_analysisGemini 3.8 Flash scores 83.15% on Terminal-Bench
artificial_analysisGemini 3.8 Flash scores 10.1% on Terminal-Bench
artificial_analysisGemini 3.8 Flash scores 84.51% on MMMU-Pro
artificial_analysisGemini 3.8 Flash scores 54.98% on SciCode
artificial_analysisGemini 3.8 Flash scores 37.07% on Humanity's Last Exam
artificial_analysisGemini 3.8 Flash scores 92.02% on GPQA Diamond
artificial_analysisGemini 3.8 Flash scores 33.79 on Artificial Analysis Intelligence Index
artificial_analysisGemini 3.8 Flash: capabilities changed from ["Caching", "Code execution", "Computer use", "File searc… to ["caching", "code_execution", "computer_use", "file_searc…
CapabilitiesCaching, Code execution, Computer use, File search, Function calling, Grounding with Google Maps, Search grounding, Structured outputs, Thinking, URL context→caching, code_execution, computer_use, file_search, function_calling, maps_grounding, reasoning, search_grounding, structured_output, url_contextgoogleGemini 3.8 Flash: context length changed from 1000000 to 1048576
Context window1M tokens→1.05M tokensgoogleGemini 3.8 Flash: family changed from Gemini to Gemini 3.8
FamilyGemini→Gemini 3.8google
Change history60
Viewing AI Atlas as of 15 Sept 2026 — attributes exactly as the atlas knew them on that day; later corrections are not shown.
Back to today →Attributes as of 15 Sept 2026 33 claims in force
- Family
- Gemini 3.8
- Version
- 3.8
- Release date
- 2 Sept 2026
- Status
- active
- Openness
- proprietary
- Context window
- 1.05M tokens
- Max output
- 65.5K tokens
- Modalities
- audio, document, image, text, video
- Input modalities
- audio, document, image, text, video
- Output modalities
- text
- Tokenizer
- Gemini
- API model id
- gemini-3.8-flash
- Aa context window
- 1,000,000
- Aa openness
- proprietary
- Audio
- Yes
- Capabilities
- caching, code_execution, computer_use, file_search, function_calling, maps_grounding, reasoning, search_grounding, structured_output, url_context
- Capabilities raw
- Caching, Code execution, Computer use, File search, Function calling, Grounding with Google Maps, Search grounding, Structured outputs, Thinking, URL context
- File input
- Yes
- Latest update
- 2026-09
- Aa median output tokens per second
- 269.2
- Openrouter id
- google/gemini-3.8-flash
- Openrouter listed at
- 2 Sept 2026
- Reasoning
- Yes
- Structured output
- Yes
- Supported parameters
- include_reasoning, max_tokens, reasoning, reasoning_effort, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_p
- Tool calling
- Yes
- Versions
- {"stable":["gemini-3.8-flash"]}
- Vision
- Yes
- Weights available
- No
Familyfamily3conflicting claims
Versionversion1
Release daterelease_date1
Statusstatus1
Opennessopenness3
Context windowcontext_length5conflicting claims
Max outputmax_output_tokens1
Modalitiesmodalities9conflicting claims
Input modalitiesmodalities_input9conflicting claims
Output modalitiesmodalities_output1
Tokenizertokenizer1
API model idapi_model_id1
Official pageofficial_url1
Aa context windowaa_context_window1
Aa opennessaa_openness1
Audioaudio1
Capabilitiescapabilities2
Capabilities rawcapabilities_raw1
Descriptiondescription1
File inputfile_input1
Latest updatelatest_update1
Aa median output tokens per secondmetric.aa_median_output_tokens_per_second1
Openrouter idopenrouter_id1
Openrouter listed atopenrouter_listed_at1
Reasoningreasoning3
Safety notessafety_notes1
Structured outputstructured_output1
Supported parameterssupported_parameters1
Tool callingtool_calling1
Training data notestraining_data_notes1
Versionsversions1
Visionvision1
Weights availableweights_available1
Claims are temporal and append-only: a new observation closes the previous claim (valid_to) instead of overwriting it. Conflicting claims from different sources are kept side by side and flagged — never averaged. Methodology →
Provenance
Attributed facts
33
Source tiers
T1T219 / 14
Freshest observation
24 h ago
Conflicts
18 flagged
Source documents 7
Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.
Data quality (73/100) measures how well AI Atlas knows this entity — completeness, primary-source ratio, freshness, conflicts — never how good the model is. Methodology →