Grok 4.3
xai/grok-4.3- xAI
- Family · Grok 4
- Released 30 Apr 2026
- Text · Image · File → Text
- Reasoning · adjustable
- Tools · JSON · Vision
Not on OSMEXO yet
Spec & list price
Context1M
Max output900K
Input / 1M$1.25
Output / 1M$2.502× input
Scores · standing
Intelligence24.9#98/489
Coding42.2#86/196
Usage
Tokens / day4.3B▲ 2.6% 7d
Usage rank#120
Requests / day604.4K
Tool calls / req0.07
Per request · 30 days
Avg input4.9K0.3× median
Avg output5970.9× median
Cache hit43%0.6× median
Reasoning share63%1.1× median
Tool-call failures0.1%0.1× median
Consumption
Peak 9.2B 26 SepLow 3.2B 1 NovRank today #120
News & trends
40 mentions this week ▲ 186%Grok 4 family · videos, posts, articles · 30 days
Trending · 14 days
Benchmarks
bar = share of the top score · Benchmark data: Artificial AnalysisSpeed · median across 3 providers, latest day
Output speed#158/33470 tok/sabout median
First token#1911.25 s1.2× longer than median
#22 cheapest of 31 with a similar score
Value
Intelligence Index
Blended price per 1M tokens (log)
Family progress
Intelligence Index
- Family releases46.4
Technical details
- Reasoning
- on · low
- Tokenizer
- Grok
- Moderation
- none
Reasoning effort
- high
- medium
- low · default
- none
Request parameters 13
- tools
- tool_choice
- response_format
- structured_outputs
- reasoning
- reasoning_effort
- include_reasoning
- temperature
- top_p
- max_tokens
+3 more
- seed
- logprobs
- top_logprobs
When to use
Best for
- Precise instruction following
- Agent workflows with tool calling
- Image and file understanding
- Long-context question answering
Think twice for
- Self-hosting (closed weights)
- Audio input
- Hardest coding agent runs
Research papers
- Assessing AI Companion Chatbot Safety for Vulnerable Users: A Crisis-Stage Multi-Platform Evaluation29 Oct
- Divide, Consult, Conquer: Capability Laundering Through Aligned LLMs14 Sep
- VDiff-Bench: A Challenging Benchmark for Fine-Grained Image Difference Identification5 Sep
- Gubernaut: A Deterministic Homeostatic Controller for Affect-Regulated LLM Agents, Validated Across Independent Model Families27 Jul
- Reliability-Contagion Feasibility in LLM Multi-Agent Networks24 Jul
- Evaluating medical AI under missing information: same-provider judges and human raters change apparent safety21 Jul
- Judge-dependent safety gains and model-specific helpfulness costs of evidence-sufficiency prompting in clinical LLMs20 Jul
- From Outliers to Errors: Auditing Pali-to-English LLM Translations with Multi-Reference Adjudication31 May
Social
Most watched and discussed · 90 days
- XAEL = Grok 4.3 : on a changé son modèle. Même IA ?AI ARENA TV · 201 views · 22 Oct
Same family
Grok 4- 25.2
- 38.8
- 46.4
- 22.5
- 44.3
Alternatives
other vendors · similar capability- 26.3
- T25.0
- 29.7
- 24.1
Want Grok 4.3 on OSMEXO?
Sign in to request it. Requests are reviewed; a request does not guarantee availability.