Gemini 3.6 Flash

google/gemini-3.6-flash-20260721

Google's Gemini 3.6 Flash is a multimodal model that accepts text, image, video, file, and audio inputs and produces text output. It supports tool usage and reasoning and can handle contexts up to 1,048,576 tokens. Pricing is set at 1.5 per million input tokens and 7.5 per million output tokens.

tool use
structured outputs
reasoning

At a glance

providergoogle

releasedJul 9, 2026

context1.0M

input / M$0.75

Specifications

What the API gives you

Context window: 1.0M tokens
Max output: 66K tokens
Input price: $0.75 / 1M tokens
Output price: $3.75 / 1M tokens
Cache read: $0.075 / 1M tokens
Modalities: text, image, video, file, audio → text
Knowledge cutoff: —
Released: Jul 9, 2026

Measured evidence

Public benchmark record

Design Arena Coding: 1313 Elo2026-07-28
Design Arena Coding: 1313 Elo2026-07-27
Design Arena Coding: 1313 Elo2026-07-30
Design Arena Coding: 1312 Elo2026-07-29
Design Arena Coding: 1314 Elo2026-07-31
Design Arena Coding: 1315 Elo2026-08-02
Design Arena Coding: 1317 Elo2026-08-05
Design Arena Coding: 1314 Elo2026-08-01
Design Arena Coding: 1318 Elo2026-08-09
Design Arena Coding: 1315 Elo2026-08-03
Design Arena Coding: 1318 Elo2026-08-04
Design Arena Coding: 1316 Elo2026-08-14
Design Arena Coding: 1317 Elo2026-08-06
Design Arena Coding: 1319 Elo2026-08-10
Design Arena Coding: 1318 Elo2026-08-07
Design Arena Coding: 1315 Elo2026-08-18
Design Arena Coding: 1317 Elo2026-08-08
Design Arena Coding: 1319 Elo2026-08-21
Design Arena Coding: 1311 Elo2026-08-15
Design Arena Coding: 1318 Elo2026-08-11
Design Arena Coding: 1318 Elo2026-08-12
Design Arena Coding: 1314 Elo2026-08-19
Design Arena Coding: 1317 Elo2026-08-13
Design Arena Coding: 1318 Elo2026-08-27
Design Arena Coding: 1314 Elo2026-08-20
Design Arena Coding: 1309 Elo2026-08-16
Design Arena Coding: 1313 Elo2026-08-17
Design Arena Coding: 1319 Elo2026-08-22
Design Arena Coding: 1319 Elo2026-08-23
Design Arena Coding: 1317 Elo2026-08-28
Design Arena Coding: 1319 Elo2026-08-24
Design Arena Coding: 1319 Elo2026-08-25
Design Arena Coding: 1319 Elo2026-08-26
Design Arena Coding: 1317 Elo2026-08-29
Design Arena Coding: 1317 Elo2026-08-30
Design Arena Coding: 1317 Elo2026-08-31
Design Arena Coding: 1317 Elo2026-09-01
Design Arena Coding: 1315 Elo2026-09-02
Design Arena Coding: 1313 Elo2026-09-03
Design Arena Coding: 1311 Elo2026-09-04
Design Arena Full-stack Agent: 1201 Elo2026-08-09
Design Arena Full-stack Agent: 1204 Elo2026-08-14
Design Arena Full-stack Agent: 1201 Elo2026-08-10
Design Arena Full-stack Agent: 1202 Elo2026-08-18
Design Arena Full-stack Agent: 1204 Elo2026-08-08
Design Arena Full-stack Agent: 1205 Elo2026-08-15
Design Arena Full-stack Agent: 1203 Elo2026-08-11
Design Arena Full-stack Agent: 1206 Elo2026-08-12
Design Arena Full-stack Agent: 1208 Elo2026-08-13
Design Arena Full-stack Agent: 1201 Elo2026-08-19
Design Arena Full-stack Agent: 1209 Elo2026-08-27
Design Arena Full-stack Agent: 1201 Elo2026-08-20
Design Arena Full-stack Agent: 1207 Elo2026-08-16
Design Arena Full-stack Agent: 1206 Elo2026-08-17
Design Arena Full-stack Agent: 1212 Elo2026-08-21
Design Arena Full-stack Agent: 1211 Elo2026-08-22
Design Arena Full-stack Agent: 1212 Elo2026-08-23
Design Arena Full-stack Agent: 1207 Elo2026-08-28
Design Arena Full-stack Agent: 1198 Elo2026-08-24
Design Arena Full-stack Agent: 1202 Elo2026-08-25
Design Arena Full-stack Agent: 1205 Elo2026-08-26
Design Arena Full-stack Agent: 1208 Elo2026-08-29
Design Arena Full-stack Agent: 1210 Elo2026-08-30
Design Arena Full-stack Agent: 1209 Elo2026-08-31
Design Arena Full-stack Agent: 1209 Elo2026-09-01
Design Arena Full-stack Agent: 1210 Elo2026-09-02
Design Arena Full-stack Agent: 1211 Elo2026-09-03
Design Arena Full-stack Agent: 1211 Elo2026-09-04
Design Arena UI: 1315 Elo2026-07-28
Design Arena UI: 1315 Elo2026-07-27
Design Arena UI: 1316 Elo2026-07-30
Design Arena UI: 1315 Elo2026-07-29
Design Arena UI: 1325 Elo2026-07-31
Design Arena UI: 1323 Elo2026-08-02
Design Arena UI: 1323 Elo2026-08-01
Design Arena UI: 1342 Elo2026-08-05
Design Arena UI: 1342 Elo2026-08-09
Design Arena UI: 1327 Elo2026-08-03
Design Arena UI: 1341 Elo2026-08-04
Design Arena UI: 1332 Elo2026-08-14
Design Arena UI: 1342 Elo2026-08-06
Design Arena UI: 1341 Elo2026-08-10
Design Arena UI: 1341 Elo2026-08-07
Design Arena UI: 1340 Elo2026-08-08
Design Arena UI: 1331 Elo2026-08-18
Design Arena UI: 1338 Elo2026-08-11
Design Arena UI: 1334 Elo2026-08-15
Design Arena UI: 1337 Elo2026-08-12
Design Arena UI: 1338 Elo2026-08-13
Design Arena UI: 1331 Elo2026-08-19
Design Arena UI: 1329 Elo2026-08-27
Design Arena UI: 1333 Elo2026-08-16
Design Arena UI: 1331 Elo2026-08-20
Design Arena UI: 1332 Elo2026-08-17
Design Arena UI: 1327 Elo2026-08-21
Design Arena UI: 1329 Elo2026-08-22
Design Arena UI: 1330 Elo2026-08-23
Design Arena UI: 1331 Elo2026-08-28
Design Arena UI: 1332 Elo2026-08-24
Design Arena UI: 1331 Elo2026-08-25
Design Arena UI: 1330 Elo2026-08-26
Design Arena UI: 1331 Elo2026-08-29
Design Arena UI: 1331 Elo2026-08-30
Design Arena UI: 1332 Elo2026-08-31
Design Arena UI: 1332 Elo2026-09-01
Design Arena UI: 1332 Elo2026-09-02
Design Arena UI: 1327 Elo2026-09-03
Design Arena UI: 1327 Elo2026-09-04
Design Arena Web-app Agent: 1252 Elo2026-08-09
Design Arena Web-app Agent: 1253 Elo2026-08-14
Design Arena Web-app Agent: 1248 Elo2026-08-10
Design Arena Web-app Agent: 1245 Elo2026-08-18
Design Arena Web-app Agent: 1250 Elo2026-08-11
Design Arena Web-app Agent: 1243 Elo2026-08-15
Design Arena Web-app Agent: 1252 Elo2026-08-12
Design Arena Web-app Agent: 1249 Elo2026-08-13
Design Arena Web-app Agent: 1245 Elo2026-08-19
Design Arena Web-app Agent: 1231 Elo2026-08-27
Design Arena Web-app Agent: 1245 Elo2026-08-20
Design Arena Web-app Agent: 1246 Elo2026-08-16
Design Arena Web-app Agent: 1248 Elo2026-08-17
Design Arena Web-app Agent: 1250 Elo2026-08-21
Design Arena Web-app Agent: 1243 Elo2026-08-22
Design Arena Web-app Agent: 1242 Elo2026-08-23
Design Arena Web-app Agent: 1231 Elo2026-08-28
Design Arena Web-app Agent: 1237 Elo2026-08-24
Design Arena Web-app Agent: 1236 Elo2026-08-25
Design Arena Web-app Agent: 1235 Elo2026-08-26
Design Arena Web-app Agent: 1230 Elo2026-08-29
Design Arena Web-app Agent: 1227 Elo2026-08-30
Design Arena Web-app Agent: 1227 Elo2026-08-31
Design Arena Web-app Agent: 1226 Elo2026-09-01
Design Arena Web-app Agent: 1226 Elo2026-09-02
Design Arena Web-app Agent: 1226 Elo2026-09-03
Design Arena Web-app Agent: 1226 Elo2026-09-04
Design Arena Website: 1321 Elo2026-07-28
Design Arena Website: 1322 Elo2026-07-30
Design Arena Website: 1319 Elo2026-07-27
Design Arena Website: 1319 Elo2026-07-29
Design Arena Website: 1322 Elo2026-07-31
Design Arena Website: 1322 Elo2026-08-02
Design Arena Website: 1322 Elo2026-08-01
Design Arena Website: 1323 Elo2026-08-05
Design Arena Website: 1322 Elo2026-08-03
Design Arena Website: 1325 Elo2026-08-09
Design Arena Website: 1323 Elo2026-08-04
Design Arena Website: 1323 Elo2026-08-14
Design Arena Website: 1322 Elo2026-08-06
Design Arena Website: 1323 Elo2026-08-07
Design Arena Website: 1324 Elo2026-08-10
Design Arena Website: 1323 Elo2026-08-08
Design Arena Website: 1317 Elo2026-08-18
Design Arena Website: 1325 Elo2026-08-11
Design Arena Website: 1314 Elo2026-08-15
Design Arena Website: 1325 Elo2026-08-12
Design Arena Website: 1325 Elo2026-08-13
Design Arena Website: 1317 Elo2026-08-19
Design Arena Website: 1314 Elo2026-08-16
Design Arena Website: 1317 Elo2026-08-20
Design Arena Website: 1323 Elo2026-08-27
Design Arena Website: 1316 Elo2026-08-17
Design Arena Website: 1323 Elo2026-08-21
Design Arena Website: 1323 Elo2026-08-22
Design Arena Website: 1323 Elo2026-08-23
Design Arena Website: 1323 Elo2026-08-24
Design Arena Website: 1323 Elo2026-08-28
Design Arena Website: 1323 Elo2026-08-25
Design Arena Website: 1323 Elo2026-08-26
Design Arena Website: 1323 Elo2026-08-29
Design Arena Website: 1322 Elo2026-08-30
Design Arena Website: 1322 Elo2026-08-31
Design Arena Website: 1321 Elo2026-09-01
Design Arena Website: 1320 Elo2026-09-02
Design Arena Website: 1318 Elo2026-09-03
Design Arena Website: 1317 Elo2026-09-04
LiveBench Agentic Coding: 43.4%2026-06-25
LiveBench Coding: 77.9%2026-06-25
LiveBench Data Analysis: 63%2026-06-25
LiveBench Instruction Following: 75.4%2026-06-25
LiveBench Language: 83.9%2026-06-25
LiveBench Math: 86.4%2026-06-25
LiveBench Reasoning: 85.2%2026-06-25
LMArena Agent: -0.62026-07-27
LMArena Agent: -1.62026-07-28
LMArena Agent: -32026-08-03
LMArena Agent: -2.42026-08-13
LMArena Agent: -2.52026-08-12
LMArena Agent: -2.72026-08-06
LMArena Agent: -2.72026-08-11
LMArena Agent: -3.22026-08-24
LMArena Agent: -3.32026-08-27
LMArena Agent: -3.52026-08-31
LMArena Creative Writing: 1468 Elo2026-07-21
LMArena Creative Writing: 1464 Elo2026-07-27
LMArena Creative Writing: 1464 Elo2026-07-30
LMArena Creative Writing: 1455 Elo2026-08-02
LMArena Creative Writing: 1455 Elo2026-08-03
LMArena Creative Writing: 1463 Elo2026-08-06
LMArena Creative Writing: 1466 Elo2026-08-12
LMArena Creative Writing: 1463 Elo2026-08-21
LMArena Creative Writing: 1462 Elo2026-08-27
LMArena Creative Writing: 1463 Elo2026-09-01
LMArena Instruction Following: 1473 Elo2026-07-21
LMArena Instruction Following: 1471 Elo2026-07-27
LMArena Instruction Following: 1471 Elo2026-07-30
LMArena Instruction Following: 1470 Elo2026-08-02
LMArena Instruction Following: 1470 Elo2026-08-03
LMArena Instruction Following: 1469 Elo2026-08-06
LMArena Instruction Following: 1470 Elo2026-08-12
LMArena Instruction Following: 1467 Elo2026-08-21
LMArena Instruction Following: 1466 Elo2026-08-27
LMArena Instruction Following: 1465 Elo2026-09-01
LMArena Long Query: 1482 Elo2026-07-21
LMArena Long Query: 1477 Elo2026-07-27
LMArena Long Query: 1477 Elo2026-07-30
LMArena Long Query: 1474 Elo2026-08-02
LMArena Long Query: 1474 Elo2026-08-03
LMArena Long Query: 1477 Elo2026-08-06
LMArena Long Query: 1476 Elo2026-08-12
LMArena Long Query: 1475 Elo2026-08-21
LMArena Long Query: 1475 Elo2026-08-27
LMArena Long Query: 1476 Elo2026-09-01
LMArena Math: 1519 Elo2026-07-21
LMArena Math: 1508 Elo2026-07-27
LMArena Math: 1508 Elo2026-07-30
LMArena Math: 1523 Elo2026-08-02
LMArena Math: 1523 Elo2026-08-03
LMArena Math: 1523 Elo2026-08-06
LMArena Math: 1514 Elo2026-08-12
LMArena Math: 1512 Elo2026-08-21
LMArena Math: 1508 Elo2026-08-27
LMArena Math: 1505 Elo2026-09-01
LMArena Text: 1491 Elo2026-07-21
LMArena Text: 1484 Elo2026-08-12
LMArena Text: 1481 Elo2026-08-21
LMArena Text: 1481 Elo2026-08-27
LMArena Text: 1480 Elo2026-09-01
LMArena Vision: 1301 Elo2026-07-26
LMArena Vision: 1301 Elo2026-07-30
LMArena Vision: 1290 Elo2026-08-02
LMArena Vision: 1295 Elo2026-08-06
LMArena Vision: 1285 Elo2026-08-21
LMArena Vision: 1285 Elo2026-08-27
LMArena WebDev: 1537 Elo2026-07-21
LMArena WebDev: 1537 Elo2026-08-12
LMArena WebDev: 1537 Elo2026-08-14
LMArena WebDev: 1539 Elo2026-08-21
LMArena WebDev: 1538 Elo2026-08-30
LMArena WebDev: 1538 Elo2026-09-01

LMArena data licensed CC-BY-4.0. Aider (Apache-2.0) and SWE-bench (MIT) are open. Design Arena uses blind human preference.

Task evidence

How broad is the support?

Each status is exact to this model version and the selected task. Official release claims remain provisional.

Best LLM for Competition MathEstablished evidence
100% of intended weighted task evidence across 2 independent sources. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Business WritingEstablished evidence
100% of intended weighted task evidence across 2 independent sources. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Logic & Problem SolvingEstablished evidence
100% of intended weighted task evidence across 2 independent sources. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Marketing CopyEstablished evidence
100% of intended weighted task evidence across 2 independent sources. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Chat & RoleplayEstablished evidence
100% of intended weighted task evidence across 2 independent sources. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Creative WritingEstablished evidence
100% of intended weighted task evidence across 2 independent sources. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Frontend & UIEstablished evidence
100% of intended weighted task evidence across 3 independent sources. 5 of 5 configured evaluation feeds currently have results.
Best LLM for WritingEstablished evidence
100% of intended weighted task evidence across 2 independent sources. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Roleplay & CharacterEstablished evidence
100% of intended weighted task evidence across 2 independent sources. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Math & ReasoningEstablished evidence
100% of intended weighted task evidence across 2 independent sources. 5 of 5 configured evaluation feeds currently have results.
Best LLM for a General AssistantEstablished evidence
100% of intended weighted task evidence across 2 independent sources. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Code CompletionEstablished evidence
81% of intended weighted task evidence across 2 independent sources. 4 of 4 configured evaluation feeds currently have results.
Best Open-Weight LLM for Self-HostingEstablished evidence
67% of intended weighted task evidence across 2 independent sources. 3 of 4 configured evaluation feeds currently have results.
Best LLM for Image UnderstandingEstablished evidence
65% of intended weighted task evidence across 2 independent sources. 3 of 4 configured evaluation feeds currently have results.
Best LLM for Fast SummariesEstablished evidence
64% of intended weighted task evidence across 2 independent sources. 3 of 4 configured evaluation feeds currently have results.
Best Value LLMEstablished evidence
64% of intended weighted task evidence across 2 independent sources. 3 of 4 configured evaluation feeds currently have results.
Best LLM for CodingProvisional evidence
58% of intended weighted task evidence across 3 independent sources; broader independent confirmation is still missing from Aider Polyglot or SWE-bench Verified. 8 of 8 configured evaluation feeds currently have results. Official release evidence is also tracked provisionally.
Best LLM for Literary TranslationProvisional evidence
55% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from WMT translation. 4 of 5 configured evaluation feeds currently have results.
Best LLM for Everyday TranslationProvisional evidence
55% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from WMT translation. 4 of 5 configured evaluation feeds currently have results.
Best LLM for TranslationProvisional evidence
55% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from WMT translation. 4 of 5 configured evaluation feeds currently have results.
Best Cheap LLM for High VolumeProvisional evidence
50% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from Route reliability. 3 of 4 configured evaluation feeds currently have results.
Cheapest Capable LLMProvisional evidence
50% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from Route reliability. 3 of 4 configured evaluation feeds currently have results.
Best Local LLM for CodingProvisional evidence
50% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from Aider Polyglot or SWE-rebench. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Agentic CodingProvisional evidence
46% of intended weighted task evidence across 3 independent sources; broader independent confirmation is still missing from Frontier-Bench or SWE-rebench. 8 of 8 configured evaluation feeds currently have results.
Best LLM for Debugging & Code ReviewProvisional evidence
40% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from Aider Polyglot or SWE-bench Verified. 6 of 6 configured evaluation feeds currently have results. Official release evidence is also tracked provisionally.
Best LLM for Document SummarizationProvisional evidence
40% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from FACTS Grounding or LMArena Document. 4 of 6 configured evaluation feeds currently have results.
Best LLM for SummarizationProvisional evidence
40% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from FACTS Grounding or LMArena Document. 4 of 6 configured evaluation feeds currently have results.
Best LLM for JSON & Schema OutputProvisional evidence
39% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or Structured-output eval. 4 of 5 configured evaluation feeds currently have results.
Best LLM for Structured Data ExtractionProvisional evidence
39% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or Structured-output eval. 4 of 5 configured evaluation feeds currently have results.
Best LLM for Hermes AgentProvisional evidence
33% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or SWE-rebench. 5 of 5 configured evaluation feeds currently have results.
Best LLM for OpenHandsProvisional evidence
33% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or SWE-rebench. 5 of 5 configured evaluation feeds currently have results.
Best LLM for GooseProvisional evidence
33% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or SWE-rebench. 5 of 5 configured evaluation feeds currently have results.
Best LLM for OpenClawProvisional evidence
33% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or SWE-rebench. 5 of 5 configured evaluation feeds currently have results.
Best LLM for High-Volume ExtractionProvisional evidence
31% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Route reliability or Structured-output eval. 2 of 4 configured evaluation feeds currently have results.
Best LLM for AI AgentsProvisional evidence
27% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or OSWorld Verified. 7 of 7 configured evaluation feeds currently have results. Official release evidence is also tracked provisionally.
Best LLM for Tool & Function CallingProvisional evidence
25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Search. 4 of 5 configured evaluation feeds currently have results. Official release evidence is also tracked provisionally.
Best LLM for Autonomous AgentsProvisional evidence
25% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from Frontier-Bench or OSWorld Verified. 7 of 7 configured evaluation feeds currently have results.
Best LLM for Massive ContextProvisional evidence
25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Document or LongBench v2. 3 of 4 configured evaluation feeds currently have results.
Best LLM for Computer & Browser UseProvisional evidence
25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or OSWorld Verified. 4 of 4 configured evaluation feeds currently have results. Official release evidence is also tracked provisionally.
Best LLM for OCR & DocumentsProvisional evidence
25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Document or Structured-output eval. 3 of 5 configured evaluation feeds currently have results.
Best LLM for Vision & DocumentsProvisional evidence
25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Document or Structured-output eval. 3 of 5 configured evaluation feeds currently have results. Official release evidence is also tracked provisionally.
Best LLM for Charts & DiagramsProvisional evidence
25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Document or Structured-output eval. 3 of 5 configured evaluation feeds currently have results. Official release evidence is also tracked provisionally.
Best LLM for Document ParsingProvisional evidence
25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Document or Structured-output eval. 3 of 5 configured evaluation feeds currently have results.
Best LLM for RAG & Long ContextProvisional evidence
18% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from FACTS Grounding or LMArena Document. 4 of 6 configured evaluation feeds currently have results. Official release evidence is also tracked provisionally.
Best LLM for Grounded RAGProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from CRAG or FACTS Grounding. 2 of 5 configured evaluation feeds currently have results.

Official release evidence · provisional

What the launch material actually says

Verbatim claims from the model provider. These help on launch day but do not count as independent confirmation.

“Customers like Hebbia and Harvey have found it particularly capable at multimodal tasks like document parsing, chart and data analysis, and report drafting.”
Google · 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber
“3.6 Flash delivers higher precision with fewer unwanted code edits and reduced execution loops, as seen in DeepSWE (49% vs. 37%), and shows significant improvement in ML Research, as seen in MLE Bench (63.9% vs. 49.7%).”
Google · 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber
“It has improved computer use capabilities as seen in OSWorld-Verified (83.0% vs. 78.4%). Computer use is now a built-in client side tool via the Gemini API and Gemini Enterprise.”
Google · 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber
“It outperforms 3.5 Flash in knowledge work, as shown by benchmarks like GDPval-AA v2 (1421 vs. 1349).”
Google · 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber

Recommendation coverage

Where this model appears

Catalog data via OpenRouter and models.dev. Last checked within the past 6 hours.