Grok 4.20

x-ai/grok-4.20-20260309

xAI: Grok 4.20, provided by x‑ai, accepts text, image, and file inputs and produces text output. It supports tools and reasoning, offers a context length of 2 000 000 tokens, and is priced at $1.25 per million input tokens and $2.5 per million output tokens. The model was released on 2026‑03‑31 and its weights are not open.

tool use
structured outputs
reasoning

At a glance

providerx-ai

releasedMar 9, 2026

context2M

input / M$1.25

Specifications

What the API gives you

Context window: 2M tokens
Max output: 1.8M tokens
Input price: $1.25 / 1M tokens
Output price: $2.50 / 1M tokens
Cache read: $0.20 / 1M tokens
Modalities: text, image, file → text
Knowledge cutoff: Sep 1, 2025
Released: Mar 9, 2026

Measured evidence

Public benchmark record

Design Arena Coding: 1231 Elo2026-07-28
Design Arena Coding: 1231 Elo2026-07-27
Design Arena Coding: 1232 Elo2026-07-30
Design Arena Coding: 1232 Elo2026-07-29
Design Arena Coding: 1232 Elo2026-07-31
Design Arena Coding: 1232 Elo2026-08-02
Design Arena Coding: 1231 Elo2026-08-05
Design Arena Coding: 1231 Elo2026-08-01
Design Arena Coding: 1231 Elo2026-08-09
Design Arena Coding: 1232 Elo2026-08-03
Design Arena Coding: 1232 Elo2026-08-04
Design Arena Coding: 1228 Elo2026-08-14
Design Arena Coding: 1232 Elo2026-08-06
Design Arena Coding: 1231 Elo2026-08-10
Design Arena Coding: 1232 Elo2026-08-07
Design Arena Coding: 1223 Elo2026-08-18
Design Arena Coding: 1231 Elo2026-08-08
Design Arena Coding: 1222 Elo2026-08-15
Design Arena Coding: 1231 Elo2026-08-11
Design Arena Coding: 1230 Elo2026-08-12
Design Arena Coding: 1229 Elo2026-08-13
Design Arena Coding: 1222 Elo2026-08-19
Design Arena Coding: 1228 Elo2026-08-27
Design Arena Coding: 1228 Elo2026-08-21
Design Arena Coding: 1222 Elo2026-08-20
Design Arena Coding: 1222 Elo2026-08-16
Design Arena Coding: 1222 Elo2026-08-17
Design Arena Coding: 1228 Elo2026-08-22
Design Arena Coding: 1229 Elo2026-08-23
Design Arena Coding: 1228 Elo2026-08-28
Design Arena Coding: 1229 Elo2026-08-24
Design Arena Coding: 1229 Elo2026-08-25
Design Arena Coding: 1229 Elo2026-08-26
Design Arena Coding: 1228 Elo2026-08-29
Design Arena Coding: 1228 Elo2026-08-30
Design Arena Coding: 1229 Elo2026-08-31
Design Arena Coding: 1229 Elo2026-09-01
Design Arena Coding: 1228 Elo2026-09-02
Design Arena Coding: 1226 Elo2026-09-03
Design Arena Coding: 1225 Elo2026-09-04
Design Arena Full-stack Agent: 1099 Elo2026-07-28
Design Arena Full-stack Agent: 1101 Elo2026-07-27
Design Arena Full-stack Agent: 1092 Elo2026-07-30
Design Arena Full-stack Agent: 1094 Elo2026-07-29
Design Arena Full-stack Agent: 1093 Elo2026-07-31
Design Arena Full-stack Agent: 1096 Elo2026-08-02
Design Arena Full-stack Agent: 1094 Elo2026-08-05
Design Arena Full-stack Agent: 1094 Elo2026-08-01
Design Arena Full-stack Agent: 1086 Elo2026-08-09
Design Arena Full-stack Agent: 1096 Elo2026-08-03
Design Arena Full-stack Agent: 1096 Elo2026-08-04
Design Arena Full-stack Agent: 1085 Elo2026-08-14
Design Arena Full-stack Agent: 1094 Elo2026-08-06
Design Arena Full-stack Agent: 1083 Elo2026-08-10
Design Arena Full-stack Agent: 1094 Elo2026-08-07
Design Arena Full-stack Agent: 1083 Elo2026-08-18
Design Arena Full-stack Agent: 1094 Elo2026-08-08
Design Arena Full-stack Agent: 1086 Elo2026-08-11
Design Arena Full-stack Agent: 1088 Elo2026-08-15
Design Arena Full-stack Agent: 1087 Elo2026-08-12
Design Arena Full-stack Agent: 1085 Elo2026-08-13
Design Arena Full-stack Agent: 1083 Elo2026-08-19
Design Arena Full-stack Agent: 1077 Elo2026-08-27
Design Arena Full-stack Agent: 1083 Elo2026-08-20
Design Arena Full-stack Agent: 1089 Elo2026-08-16
Design Arena Full-stack Agent: 1088 Elo2026-08-17
Design Arena Full-stack Agent: 1088 Elo2026-08-21
Design Arena Full-stack Agent: 1087 Elo2026-08-22
Design Arena Full-stack Agent: 1086 Elo2026-08-23
Design Arena Full-stack Agent: 1077 Elo2026-08-28
Design Arena Full-stack Agent: 1074 Elo2026-08-24
Design Arena Full-stack Agent: 1078 Elo2026-08-25
Design Arena Full-stack Agent: 1078 Elo2026-08-26
Design Arena Full-stack Agent: 1077 Elo2026-08-29
Design Arena Full-stack Agent: 1077 Elo2026-08-30
Design Arena Full-stack Agent: 1077 Elo2026-08-31
Design Arena Full-stack Agent: 1078 Elo2026-09-01
Design Arena Full-stack Agent: 1078 Elo2026-09-02
Design Arena Full-stack Agent: 1078 Elo2026-09-03
Design Arena Full-stack Agent: 1078 Elo2026-09-04
Design Arena UI: 1218 Elo2026-07-28
Design Arena UI: 1219 Elo2026-07-27
Design Arena UI: 1219 Elo2026-07-30
Design Arena UI: 1219 Elo2026-07-29
Design Arena UI: 1219 Elo2026-07-31
Design Arena UI: 1220 Elo2026-08-02
Design Arena UI: 1220 Elo2026-08-01
Design Arena UI: 1220 Elo2026-08-05
Design Arena UI: 1217 Elo2026-08-09
Design Arena UI: 1221 Elo2026-08-03
Design Arena UI: 1221 Elo2026-08-04
Design Arena UI: 1216 Elo2026-08-14
Design Arena UI: 1219 Elo2026-08-06
Design Arena UI: 1219 Elo2026-08-07
Design Arena UI: 1217 Elo2026-08-10
Design Arena UI: 1217 Elo2026-08-08
Design Arena UI: 1215 Elo2026-08-18
Design Arena UI: 1218 Elo2026-08-11
Design Arena UI: 1215 Elo2026-08-15
Design Arena UI: 1218 Elo2026-08-12
Design Arena UI: 1218 Elo2026-08-13
Design Arena UI: 1215 Elo2026-08-19
Design Arena UI: 1214 Elo2026-08-16
Design Arena UI: 1215 Elo2026-08-20
Design Arena UI: 1212 Elo2026-08-27
Design Arena UI: 1214 Elo2026-08-17
Design Arena UI: 1213 Elo2026-08-21
Design Arena UI: 1213 Elo2026-08-22
Design Arena UI: 1213 Elo2026-08-23
Design Arena UI: 1213 Elo2026-08-24
Design Arena UI: 1212 Elo2026-08-28
Design Arena UI: 1213 Elo2026-08-25
Design Arena UI: 1213 Elo2026-08-26
Design Arena UI: 1213 Elo2026-08-29
Design Arena UI: 1213 Elo2026-08-30
Design Arena UI: 1213 Elo2026-08-31
Design Arena UI: 1213 Elo2026-09-01
Design Arena UI: 1213 Elo2026-09-02
Design Arena UI: 1211 Elo2026-09-03
Design Arena UI: 1209 Elo2026-09-04
Design Arena Web-app Agent: 1197 Elo2026-07-28
Design Arena Web-app Agent: 1197 Elo2026-07-27
Design Arena Web-app Agent: 1196 Elo2026-07-30
Design Arena Web-app Agent: 1196 Elo2026-07-29
Design Arena Web-app Agent: 1196 Elo2026-07-31
Design Arena Web-app Agent: 1197 Elo2026-08-02
Design Arena Web-app Agent: 1197 Elo2026-08-01
Design Arena Web-app Agent: 1192 Elo2026-08-05
Design Arena Web-app Agent: 1186 Elo2026-08-09
Design Arena Web-app Agent: 1197 Elo2026-08-03
Design Arena Web-app Agent: 1197 Elo2026-08-04
Design Arena Web-app Agent: 1181 Elo2026-08-14
Design Arena Web-app Agent: 1192 Elo2026-08-06
Design Arena Web-app Agent: 1183 Elo2026-08-10
Design Arena Web-app Agent: 1192 Elo2026-08-07
Design Arena Web-app Agent: 1180 Elo2026-08-18
Design Arena Web-app Agent: 1192 Elo2026-08-08
Design Arena Web-app Agent: 1183 Elo2026-08-11
Design Arena Web-app Agent: 1181 Elo2026-08-15
Design Arena Web-app Agent: 1183 Elo2026-08-12
Design Arena Web-app Agent: 1182 Elo2026-08-13
Design Arena Web-app Agent: 1180 Elo2026-08-19
Design Arena Web-app Agent: 1175 Elo2026-08-27
Design Arena Web-app Agent: 1180 Elo2026-08-20
Design Arena Web-app Agent: 1183 Elo2026-08-16
Design Arena Web-app Agent: 1183 Elo2026-08-17
Design Arena Web-app Agent: 1180 Elo2026-08-21
Design Arena Web-app Agent: 1181 Elo2026-08-22
Design Arena Web-app Agent: 1181 Elo2026-08-23
Design Arena Web-app Agent: 1176 Elo2026-08-28
Design Arena Web-app Agent: 1175 Elo2026-08-24
Design Arena Web-app Agent: 1176 Elo2026-08-25
Design Arena Web-app Agent: 1177 Elo2026-08-26
Design Arena Web-app Agent: 1175 Elo2026-08-29
Design Arena Web-app Agent: 1175 Elo2026-08-30
Design Arena Web-app Agent: 1174 Elo2026-08-31
Design Arena Web-app Agent: 1174 Elo2026-09-01
Design Arena Web-app Agent: 1174 Elo2026-09-02
Design Arena Web-app Agent: 1174 Elo2026-09-03
Design Arena Web-app Agent: 1174 Elo2026-09-04
Design Arena Website: 1238 Elo2026-07-28
Design Arena Website: 1238 Elo2026-07-30
Design Arena Website: 1237 Elo2026-07-27
Design Arena Website: 1238 Elo2026-07-29
Design Arena Website: 1238 Elo2026-07-31
Design Arena Website: 1238 Elo2026-08-02
Design Arena Website: 1238 Elo2026-08-01
Design Arena Website: 1238 Elo2026-08-05
Design Arena Website: 1238 Elo2026-08-03
Design Arena Website: 1238 Elo2026-08-09
Design Arena Website: 1238 Elo2026-08-04
Design Arena Website: 1236 Elo2026-08-14
Design Arena Website: 1238 Elo2026-08-06
Design Arena Website: 1239 Elo2026-08-07
Design Arena Website: 1237 Elo2026-08-10
Design Arena Website: 1237 Elo2026-08-08
Design Arena Website: 1228 Elo2026-08-18
Design Arena Website: 1238 Elo2026-08-11
Design Arena Website: 1227 Elo2026-08-15
Design Arena Website: 1237 Elo2026-08-12
Design Arena Website: 1236 Elo2026-08-13
Design Arena Website: 1228 Elo2026-08-19
Design Arena Website: 1227 Elo2026-08-16
Design Arena Website: 1228 Elo2026-08-20
Design Arena Website: 1237 Elo2026-08-27
Design Arena Website: 1227 Elo2026-08-17
Design Arena Website: 1236 Elo2026-08-21
Design Arena Website: 1237 Elo2026-08-22
Design Arena Website: 1237 Elo2026-08-23
Design Arena Website: 1237 Elo2026-08-24
Design Arena Website: 1237 Elo2026-08-28
Design Arena Website: 1237 Elo2026-08-25
Design Arena Website: 1238 Elo2026-08-26
Design Arena Website: 1237 Elo2026-08-29
Design Arena Website: 1237 Elo2026-08-30
Design Arena Website: 1237 Elo2026-08-31
Design Arena Website: 1238 Elo2026-09-01
Design Arena Website: 1237 Elo2026-09-02
Design Arena Website: 1235 Elo2026-09-03
Design Arena Website: 1235 Elo2026-09-04
LMArena WebDev: 1381 Elo2026-07-21
LMArena WebDev: 1374 Elo2026-07-27
LMArena WebDev: 1375 Elo2026-07-24
LMArena WebDev: 1374 Elo2026-07-28
LMArena WebDev: 1374 Elo2026-07-30
LMArena WebDev: 1374 Elo2026-08-01
LMArena WebDev: 1374 Elo2026-07-31
LMArena WebDev: 1374 Elo2026-08-06
LMArena WebDev: 1374 Elo2026-08-02
LMArena WebDev: 1374 Elo2026-08-12
LMArena WebDev: 1374 Elo2026-08-10
LMArena WebDev: 1374 Elo2026-08-14
LMArena WebDev: 1374 Elo2026-08-21
LMArena WebDev: 1374 Elo2026-08-30
LMArena WebDev: 1373 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 Frontend & UIEstablished evidence
90% of intended weighted task evidence across 2 independent sources. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Code CompletionProvisional evidence
44% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Aider Polyglot or LiveBench Coding. 4 of 4 configured evaluation feeds currently have results.
Best LLM for CodingProvisional evidence
42% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from Aider Polyglot or LiveBench Coding. 8 of 8 configured evaluation feeds currently have results.
Best Local LLM for CodingProvisional evidence
26% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Aider Polyglot or LiveBench Coding. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Debugging & Code ReviewProvisional evidence
25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Aider Polyglot or LiveBench Coding. 6 of 6 configured evaluation feeds currently have results.
Best LLM for Agentic CodingProvisional evidence
23% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Frontier-Bench or LiveBench Agentic Coding. 8 of 8 configured evaluation feeds currently have results.
Best Cheap LLM for High VolumeProvisional evidence
17% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LMArena Text. 3 of 4 configured evaluation feeds currently have results.
Cheapest Capable LLMProvisional evidence
17% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LMArena Text. 3 of 4 configured evaluation feeds currently have results.
Best LLM for Fast SummariesProvisional evidence
14% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LMArena Text. 3 of 4 configured evaluation feeds currently have results.
Best Value LLMProvisional evidence
14% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LMArena Text. 3 of 4 configured evaluation feeds currently have results.
Best LLM for High-Volume ExtractionProvisional evidence
13% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Data Analysis or Route reliability. 2 of 4 configured evaluation feeds currently have results.
Best LLM for Literary TranslationProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Language. 4 of 5 configured evaluation feeds currently have results.
Best LLM for Everyday TranslationProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Language. 4 of 5 configured evaluation feeds currently have results.
Best LLM for TranslationProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Language. 4 of 5 configured evaluation feeds currently have results.
Best LLM for Document SummarizationProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from FACTS Grounding or LiveBench Instruction Following. 4 of 6 configured evaluation feeds currently have results.
Best LLM for Tool & Function CallingProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 4 of 5 configured evaluation feeds currently have results.
Best LLM for Business WritingProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LMArena Creative Writing. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Logic & Problem SolvingProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Reasoning. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Marketing CopyProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LMArena Creative Writing. 5 of 5 configured evaluation feeds currently have results.
Best LLM for SummarizationProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from FACTS Grounding or LiveBench Instruction Following. 4 of 6 configured evaluation feeds currently have results.
Best LLM for Chat & RoleplayProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Language. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Creative WritingProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Language. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Computer & Browser UseProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 4 of 4 configured evaluation feeds currently have results.
Best LLM for OCR & DocumentsProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LMArena Document or LMArena Vision. 3 of 5 configured evaluation feeds currently have results.
Best LLM for Image UnderstandingProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Reasoning or LMArena Vision. 3 of 4 configured evaluation feeds currently have results.
Best LLM for Vision & DocumentsProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LMArena Document or LMArena Vision. 3 of 5 configured evaluation feeds currently have results.
Best LLM for WritingProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Language. 5 of 5 configured evaluation feeds currently have results.
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.
Best LLM for Roleplay & CharacterProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Language. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Charts & DiagramsProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LMArena Document or LMArena Vision. 3 of 5 configured evaluation feeds currently have results.
Best LLM for Document ParsingProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LMArena Document or LMArena Vision. 3 of 5 configured evaluation feeds currently have results.
Best LLM for a General AssistantProvisional evidence
10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Reasoning. 5 of 5 configured evaluation feeds currently have results.
Best Open-Weight LLM for Self-HostingProvisional evidence
8% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Instruction Following or LMArena Text. 3 of 4 configured evaluation feeds currently have results.
Best LLM for Massive ContextProvisional evidence
6% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LMArena Document or LMArena Long Query. 3 of 4 configured evaluation feeds currently have results.
Best LLM for Hermes AgentProvisional evidence
6% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 5 of 5 configured evaluation feeds currently have results.
Best LLM for OpenHandsProvisional evidence
6% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 5 of 5 configured evaluation feeds currently have results.
Best LLM for GooseProvisional evidence
6% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 5 of 5 configured evaluation feeds currently have results.
Best LLM for RAG & Long ContextProvisional evidence
6% of intended weighted task evidence across 0 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 OpenClawProvisional evidence
6% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 5 of 5 configured evaluation feeds currently have results.
Best LLM for JSON & Schema OutputProvisional evidence
6% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LiveBench Data Analysis. 4 of 5 configured evaluation feeds currently have results.
Best LLM for Structured Data ExtractionProvisional evidence
6% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LiveBench Data Analysis. 4 of 5 configured evaluation feeds currently have results.
Best LLM for Competition MathProvisional evidence
5% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Mathematics or LiveBench Reasoning. 5 of 5 configured evaluation feeds currently have results.
Best LLM for Autonomous AgentsProvisional evidence
5% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Frontier-Bench or LiveBench Agentic Coding. 7 of 7 configured evaluation feeds currently have results.
Best LLM for Math & ReasoningProvisional evidence
5% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Mathematics or LiveBench Reasoning. 5 of 5 configured evaluation feeds currently have results.
Best LLM for AI AgentsProvisional evidence
5% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LiveBench Agentic Coding. 7 of 7 configured evaluation feeds currently have results.

Recommendation coverage

Where this model appears

Not currently included in a published recommendation. The task evidence panel above says exactly what is confirmed and what is still missing.

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