modelbench日本語

Gemini 3.7 Flash

GoogleLegacy

Input price
—
Per 1 million tokens (USD)
Output price
—
Per 1 million tokens (USD)
Effort levels
3total
Default not stated
Benchmark results
22total
Benchmark types: 22

Model information

API model ID
gemini-3.7-flash
release date
2026-08-13
effort
Low / Med / High
Effort source
Based on Thinking: Supported (low, medium, high) and minimal not supported on the Gemini API model page (aistudio.google.com/docs/models/gemini-3.7-flash). The default thinking level is not stated. Discontinued; gemini-3.7-flash in the API automatically redirects to gemini-3.8-flash.
Claude Code
unavailable
Codex
unavailable
Availability notes
Both are false because the model appears in neither the Claude Code official model list (code.claude.com/docs/en/model-config alias table, retrieved 2026-11-13; original: the table for "Use a model alias to select model settings without remembering exact version numbers" contains only Anthropic models such as fable, sonnet, opus, and haiku) nor the Codex official model list (learn.chatgpt.com/docs/models, retrieved 2026-11-13; original: "Recommended models" contains only the GPT-6 series) (connections via a custom provider are not counted as official options, and connections via a gateway are not counted as official options).
Availability source
https://code.claude.com/docs/en/model-config

Price (USD per 1 million tokens)

Prices do not vary by effort; token consumption does.

input—
output—
cache read—
cache write—

The model card table lists input $0.75 and output $3.75 (introductory price, $1.50 / $7.50 from 2027-01-01), but no unit price is entered because it has been discontinued and the API is redirected to gemini-3.8-flash.

Benchmark results

BenchmarkeffortScoreSource cost (metric)measured/reported bysourceConditions and extraction method
Terminal-Bench 3.0not statednot stated14.9%—GoogleGemini 3.7 Flash model cardofficial · Retrieved 2026-11-13Vendor-reported value. Table in the Google model card. methodology PDF (deepmind.com/models/evals-methodology/gemini-3-7-flash): All of the results are run with the Gemini API ... with default sampling settings unless indicated otherwise. Gemini computed in-house; mini swe agent harness (LiteLLM 1.96). Minor changes to 2 tasks for compatibility. thinking level not statedtranscribed from table
Source excerptTerminal-bench 3.0 General agent capabilities | 14.9%Location Gemini 3.7 Flash model card Evaluation table, Terminal-bench 3.0 row, Gemini 3.7 Flash column. In-house computation basis: methodology PDF, Coding / Terminal-Bench 3
Terminal-Bench 2.1not statednot stated85.8%—GoogleGemini 3.7 Flash model cardofficial · Retrieved 2026-11-13Vendor-reported value. Table in the Google model card. methodology PDF (deepmind.com/models/evals-methodology/gemini-3-7-flash): All of the results are run with the Gemini API ... with default sampling settings unless indicated otherwise. Gemini computed in-house; default harness Terminus 2. thinking level not statedtranscribed from table
Source excerptTerminal-bench 2.1 Agentic terminal coding | 85.8%Location Gemini 3.7 Flash model card Evaluation table, Terminal-bench 2.1 row, Gemini 3.7 Flash column. In-house computation basis: methodology PDF, Coding / Terminal-Bench 2.1
DeepSWE v1.1HighHigh65.3%—GoogleGemini 3.7 Flash model cardofficial · Retrieved 2026-11-13Vendor-reported value. Table in the Google model card. methodology PDF (deepmind.com/models/evals-methodology/gemini-3-7-flash): All of the results are run with the Gemini API ... with default sampling settings unless indicated otherwise. Results for Gemini 3.7 Flash are self computed, and use a mini swe agent harness with LiteLLM version 1.96 and high thinking (3.6 Flash is not used because its value is from the Datacurve leaderboard)transcribed from table
Source excerptDeepSWE v1.1 Long-horizon software engineering | 65.3%Location Gemini 3.7 Flash model card Evaluation table, DeepSWE v1.1 row, Gemini 3.7 Flash column. In-house computation basis: methodology PDF, Coding / DeepSWE v1.1
HLE-Verifiednot statednot stated53.6%—GoogleGemini 3.7 Flash model cardofficial · Retrieved 2026-11-13Vendor-reported value. Table in the Google model card. methodology PDF (deepmind.com/models/evals-methodology/gemini-3-7-flash): All of the results are run with the Gemini API ... with default sampling settings unless indicated otherwise. All models computed in-house; 1,811 verified questionstranscribed from table
Source excerptHLE-Verified Multidisciplinary expert reasoning | 53.6%Location Gemini 3.7 Flash model card Evaluation table, HLE-Verified row, Gemini 3.7 Flash column. In-house computation basis: methodology PDF, Scientific Reasoning / HLE-Verified
OSWorld 2.0not statednot stated47.9%—GoogleGemini 3.7 Flash model cardofficial · Retrieved 2026-11-13Vendor-reported value. Table in the Google model card. methodology PDF (deepmind.com/models/evals-methodology/gemini-3-7-flash): All of the results are run with the Gemini API ... with default sampling settings unless indicated otherwise. Computed in-house; partial score; maximum of 3 runs (1 trial each); up to 500 steps; run before the official 08.08 patchtranscribed from table
Source excerptOSWorld-2.0 Agentic computer use | 47.9%Location Gemini 3.7 Flash model card Evaluation table, OSWorld-2.0 row, Gemini 3.7 Flash column. In-house computation basis: methodology PDF, Computer use / OSWorld 2.0
Agents' Last Exam (pass rate, version not stated)not statednot stated26.3%—GoogleGemini 3.7 Flash model cardofficial · Retrieved 2026-11-13Vendor-reported value. Table in the Google model card. methodology PDF (deepmind.com/models/evals-methodology/gemini-3-7-flash): All of the results are run with the Gemini API ... with default sampling settings unless indicated otherwise. Computed in-house; default ALE-Claw harness. version not statedtranscribed from table
Source excerptAgent's Last Exam Multimodal desktop and OS agent tasks | Pass rate | 26.3%Location Gemini 3.7 Flash model card Evaluation table, Agent's Last Exam(Pass rate) row, Gemini 3.7 Flash column. In-house computation basis: methodology PDF, Computer use / Agent's Last Exam
LABBench 2not statednot stated82.1%—GoogleGemini 3.7 Flash model cardofficial · Retrieved 2026-11-13Vendor-reported value. Table in the Google model card. methodology PDF (deepmind.com/models/evals-methodology/gemini-3-7-flash): All of the results are run with the Gemini API ... with default sampling settings unless indicated otherwise. Gemini computed in-housetranscribed from table
Source excerptLABBench2 Biology real-world research tasks | 82.1%Location Gemini 3.7 Flash model card Evaluation table, LABBench2 row, Gemini 3.7 Flash column. In-house computation basis: methodology PDF, Scientific Reasoning / LABBench2
LVBenchnot statednot stated85.4%—GoogleGemini 3.7 Flash model cardofficial · Retrieved 2026-11-13Vendor-reported value. Table in the Google model card. methodology PDF (deepmind.com/models/evals-methodology/gemini-3-7-flash): All of the results are run with the Gemini API ... with default sampling settings unless indicated otherwise. Computed in-house; 1024 frames; without toolstranscribed from table
Source excerptLVBench Long video understanding | 85.4%Location Gemini 3.7 Flash model card Evaluation table, LVBench row, Gemini 3.7 Flash column. In-house computation basis: methodology PDF, Multimodal / LVBench
GDP.PDFnot statednot stated34.0%—GoogleGemini 3.7 Flash model cardofficial · Retrieved 2026-11-13Vendor-reported value. Table in the Google model card. methodology PDF (deepmind.com/models/evals-methodology/gemini-3-7-flash): All of the results are run with the Gemini API ... with default sampling settings unless indicated otherwise. Gemini computed in-house. Metric name not statedtranscribed from table
Source excerptGDP.pdf Expert PDF document comprehension | 34.0%Location Gemini 3.7 Flash model card Evaluation table, GDP.pdf row, Gemini 3.7 Flash column. In-house computation basis: methodology PDF, Multimodal / GDP.PDF
CharXiv Reasoningnot statednot stated84.5%—GoogleGemini 3.7 Flash model cardofficial · Retrieved 2026-11-13Vendor-reported value. Table in the Google model card. methodology PDF (deepmind.com/models/evals-methodology/gemini-3-7-flash): All of the results are run with the Gemini API ... with default sampling settings unless indicated otherwise. Gemini computed in-house; without toolstranscribed from table
Source excerptCharXiv Reasoning Information synthesis from complex charts | No tools | 84.5%Location Gemini 3.7 Flash model card Evaluation table, CharXiv Reasoning(No tools) row, Gemini 3.7 Flash column. In-house computation basis: methodology PDF, Multimodal / CharXiv Reasoning
BioMysteryBench(Human Solvable)not statednot stated87.1%—GoogleGemini 3.7 Flash model cardofficial · Retrieved 2026-11-13Vendor-reported value. Table in the Google model card. methodology PDF (deepmind.com/models/evals-methodology/gemini-3-7-flash): All of the results are run with the Gemini API ... with default sampling settings unless indicated otherwise. Gemini computed in-house; Linux terminal and internet restricted to allowed domainstranscribed from table
Source excerptBioMysteryBench Bioinformatics research reasoning | Human solvable | 87.1%Location Gemini 3.7 Flash model card Evaluation table, BioMysteryBench(Human solvable) row, Gemini 3.7 Flash column. In-house computation basis: methodology PDF, Scientific Reasoning / BioMysteryBench
BioMysteryBench(Human Difficult)not statednot stated43.5%—GoogleGemini 3.7 Flash model cardofficial · Retrieved 2026-11-13Vendor-reported value. Table in the Google model card. methodology PDF (deepmind.com/models/evals-methodology/gemini-3-7-flash): All of the results are run with the Gemini API ... with default sampling settings unless indicated otherwise. Gemini computed in-housetranscribed from table
Source excerptHuman difficult | 43.5%Location Gemini 3.7 Flash model card Evaluation table, BioMysteryBench(Human difficult) row, Gemini 3.7 Flash column. In-house computation basis: methodology PDF, Scientific Reasoning / BioMysteryBench
GPQA DiamondHighHigh94.8%—Epoch AIEpoch AI, Capabilities & benchmarking (benchmark_data.zip, gpqa_diamond.csv) CC BY 4.0third-party · Retrieved 2026-11-12Run by Epoch AI (third-party measurement). Source label "Gemini 3.7 Flash (high)". effort is indicated by the suffix _high in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table
Source excerptgpqa_diamond.csv: Model version=gemini-3.7-flash_high, Best score (across scorers)=0.9482323232323232, stderr=0.013456525292246259, Started at=2026-08-14T18:29:12.000ZLocation gpqa_diamond.csv, data row 72 (file line 125), id=XyVL3VffJ6UboY97vxpohL, Model version=gemini-3.7-flash_high
FrontierMath Tiers 1-3(v2)HighHigh71.6%—Epoch AIEpoch AI, Capabilities & benchmarking (benchmark_data.zip, frontiermath_tiers_1_3_v2.csv) CC BY 4.0third-party · Retrieved 2026-11-12Run by Epoch AI (third-party measurement). Source label "Gemini 3.7 Flash (high)". effort is indicated by the suffix _high in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table
Source excerptfrontiermath_tiers_1_3_v2.csv: Model version=gemini-3.7-flash_high, Best score (across scorers)=0.7157894736842105, stderr=0.026764156649364653, Started at=2026-08-14T19:59:58.000ZLocation frontiermath_tiers_1_3_v2.csv, data row 78 (file line 104), id=axMdgNDC8UYYLnZYLHZpF5, Model version=gemini-3.7-flash_high
FrontierMath Tier 4(v2)HighHigh36.6%—Epoch AIEpoch AI, Capabilities & benchmarking (benchmark_data.zip, frontiermath_tier_4_v2.csv) CC BY 4.0third-party · Retrieved 2026-11-12Run by Epoch AI (third-party measurement). Source label "Gemini 3.7 Flash (high)". effort is indicated by the suffix _high in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table
Source excerptfrontiermath_tier_4_v2.csv: Model version=gemini-3.7-flash_high, Best score (across scorers)=0.36585365853658536, stderr=0.07615851217559022, Started at=2026-08-14T19:59:58.000ZLocation frontiermath_tier_4_v2.csv, data row 32 (file line 65), id=KptKXK5nE2Ns2WbiaSSnbc, Model version=gemini-3.7-flash_high
OTIS Mock AIME 2024-2025HighHigh97.2%—Epoch AIEpoch AI, Capabilities & benchmarking (benchmark_data.zip, otis_mock_aime_2024_2025.csv) CC BY 4.0third-party · Retrieved 2026-11-12Run by Epoch AI (third-party measurement). Source label "Gemini 3.7 Flash (high)". effort is indicated by the suffix _high in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table
Source excerptotis_mock_aime_2024_2025.csv: Model version=gemini-3.7-flash_high, Best score (across scorers)=0.9722222222222222, stderr=0.018586784946171872, Started at=2026-08-14T18:29:12.000ZLocation otis_mock_aime_2024_2025.csv, data row 80 (file line 140), id=kQjrYCaa7yYyMXofeR23Dd, Model version=gemini-3.7-flash_high
SimpleQA VerifiedHighHigh69.2%—Epoch AIEpoch AI, Capabilities & benchmarking (benchmark_data.zip, simpleqa_verified.csv) CC BY 4.0third-party · Retrieved 2026-11-12Run by Epoch AI (third-party measurement). Source label "Gemini 3.7 Flash (high)". effort is indicated by the suffix _high in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table
Source excerptsimpleqa_verified.csv: Model version=gemini-3.7-flash_high, Best score (across scorers)=0.692, stderr=0.01460648312734278, Started at=2026-08-27T19:30:25.000ZLocation simpleqa_verified.csv, data row 58 (file line 103), id=84ARuGyhQsjgh8umRmBvas, Model version=gemini-3.7-flash_high
Chess Puzzles(Epoch AI)HighHigh66.0%—Epoch AIEpoch AI, Capabilities & benchmarking (benchmark_data.zip, chess_puzzles.csv) CC BY 4.0third-party · Retrieved 2026-11-12Run by Epoch AI (third-party measurement). Source label "Gemini 3.7 Flash (high)". effort is indicated by the suffix _high in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table
Source excerptchess_puzzles.csv: Model version=gemini-3.7-flash_high, Best score (across scorers)=0.66, stderr=0.04760952285695234, Started at=2026-10-13T13:00:23.000ZLocation chess_puzzles.csv, data row 154 (file line 210), id=Xu34dKTAdpBWL6advf6DMg, Model version=gemini-3.7-flash_high
MirrorCodeHighHigh41.1%—Epoch AIEpoch AI, Capabilities & benchmarking (benchmark_data.zip, mirrorcode.csv) CC BY 4.0third-party · Retrieved 2026-11-12Run by Epoch AI (third-party measurement). Source label "Gemini 3.7 Flash (high)". effort is indicated by the suffix _high in Epoch's Model version. Data retrieved 2026-11-12.transcribed from table
Source excerptmirrorcode.csv: Model version=gemini-3.7-flash_high, Best score (across scorers)=0.4111111111111111, stderr=0.12157113738962926, Started at=2026-10-21T16:49:08.000ZLocation mirrorcode.csv, data row 12 (file line 17), id=eeq6feqZLzfjfVsx4aeAWL, Model version=gemini-3.7-flash_high
Epoch Capabilities Index(ECI)not statednot stated158.36—Epoch AIEpoch AI, epoch_capabilities_index/eci_scores.csv (CC BY 4.0)third-party · Retrieved 2026-11-12A per-model metric that combines results across multiple benchmarks and multiple efforts (not per effort). Calculated by Epoch AI.transcribed from table
Source excerpteci_scores.csv: Model=Gemini 3.7 Flash, eci=158.36, eci_ci_low=156.35, eci_ci_high=160.59Location epoch_capabilities_index/eci_scores.csv, data row 25 (file line 26), Model=Gemini 3.7 Flash
CharXiv Reasoning (with tools)not statednot stated88.7%—GoogleGemini 3.7 Flash model cardofficial · Retrieved 2026-11-13Vendor-reported value. Table in the Google model card. methodology PDF (deepmind.com/models/evals-methodology/gemini-3-7-flash): All of the results are run with the Gemini API ... with default sampling settings unless indicated otherwise. Gemini computed in-house; tools are search and code executiontranscribed from table
Source excerptWith tools | 88.7%Location Gemini 3.7 Flash model card Evaluation table, CharXiv Reasoning(With tools) row, Gemini 3.7 Flash column. In-house computation basis: methodology PDF, Multimodal / CharXiv Reasoning
MRCR v2 (8-needle, 128k)not statednot stated97.0%—GoogleGemini 3.7 Flash model cardofficial · Retrieved 2026-11-13Vendor-reported value. Table in the Google model card. methodology PDF (deepmind.com/models/evals-methodology/gemini-3-7-flash): All of the results are run with the Gemini API ... with default sampling settings unless indicated otherwise. Computed in-house; cumulative score up to 128ktranscribed from table
Source excerptGDM-MRCR v2 (8-needle) Long context performance | 128k (average) | 97.0%Location Gemini 3.7 Flash model card Evaluation table, GDM-MRCR v2 (8-needle) 128k (average) row, Gemini 3.7 Flash column. In-house computation basis: methodology PDF, Long Context / GDM-MRCR v2

Related rankings

References