whatmodel.sh / Grok 4.20 Multi-Agent (0309)
Grok 4.20 Multi-Agent (0309)
↗ share
+ contribute
Grok 4.20 Multi-Agent (0309) is a closed-weights language model from SpaceXAI (xAI), released March 2026.
Catalogue identity notes
Arena entry 'grok-4.20-multi-agent-beta-0309' (text, vision, search). Released 2026-03: xAI API release notes, March 2026: 'Grok 4.20 and Grok 4.20 Multi-agent are live'. xAI model page 'Grok 4.20 Multi-Agent Beta': model name grok-4.20-multi-agent-0309, aliases include grok-4.20-multi-agent-beta-0309 (the Arena entry name). A multi-agent system served as one API model; distinct from grok-4-20 (grok-4.20-0309-reasoning).
$ how it compares
The LLMs ranked either side of Grok 4.20 Multi-Agent (0309), out of 828 with a score, then more from SpaceXAI (xAI).
| # | model | score | where | ||
|---|---|---|---|---|---|
| 58 |  Dola Seed 2.0 Pro | 74±2 | comparison only | ||
| 59 |  Qwen3.8 27B | 73±1 | api $1.60/M +local | ||
| 60 | Grok 4.20 Multi-Agent (0309) | 73±2 | api $1.56/M | ||
| 61 |  Kimi K2.6 | 73±1 | api $1.71/M +local | ||
| 62 |  GPT-5 Pro | 73±1 | api $41.25/M | ||
| more from SpaceXAI (xAI) | |||||
| 30 | Grok 4.6 | 77±1 | api $3.00/M | ||
| 42 | Grok 4.5 | 75±1 | api $3.00/M | ||
| 47 | Grok 4.7 | 75±1 | api $3.00/M | ||
$ what it's good at
chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #33 of 393 * · 47% of its score · 1 test
Beats Claude Sonnet 5.5, GPT-6 Sol, GPT-5.6 Terra here, though they rank higher overall. Trails Grok 4.20 (0309 reasoning), Claude Opus 4.5, Gemini 3 Flash (Preview) here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| LMArena (text) | 1470.6 ✎ | ▁▁▁▂▃▃▃▆▆▇█▃ beats 341/379 | 47% |
vision & documents▁▁▃▂▁▃▄▂▆▇█▄ #45 of 147 * · 32% of its score · 1 test
Beats MiMo-V2.6-Pro, MiMo-V2.6-Flash (309B-A15B), DeepSeek V4 Flash Vision (Exp) here, though they rank higher overall. Trails Qwen3.6 Flash, Gemini 3.5 Flash-Lite, Kimi K2.5 here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Arena vision (style controlled) | 1250.3 ✎ | ▁▁▂▂▂▂▄▅▅▆██ beats 102/145 | 32% |
web search▂▁▂▆▄█▃▃▁▁·▁ #25 of 80 * · 19% of its score · 1 test
Beats Claude Opus 4.8, Claude Haiku 5.5, Gemini 3.1 Pro (Preview) here, though they rank higher overall. Trails GPT-4o mini (2024-07-18), Mistral Small 3.1 24B (2503), GPT-4o (2024-08-06) here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Arena search systems (style controlled) | 1204.9 ✎ | ▁·▁▁▂▂▂█▅▂▁▁ beats 24/30 | 19% |
$ where to run it
Through an API
api · xAI · $1.56/M · $1.25 in / $2.5 out per 1M tokens · 1M context · docs.x.ai ↗ · price sourceUSD per 1M tokens, xAI API text pricing table, grok-4.20-multi-agent-0309 (< 200k prompt tokens): input $1.25, output $2.50; cached input $0.20 and the ≥200k long-context tier ($2.50/$5.00) excluded. Currency: USD. xAI official models and pricing page, observed 2026-09-28.
$ how the score was worked out · LMArena leaderboard dataset 98%
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 LMArena leaderboard dataset | 3 | 98% |
| prior · pull toward the middle | 2% | |
| weighted average | = 73.1 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json