Scores on standardized evaluations and this model’s own outputs on shared benchmark prompts. Higher percentages are better — and each rank shows where this model lands among the models evaluated on that source.
| Source | Benchmark | Score |
|---|---|---|
| Artificial Analysis | Grok 4.6 (high) Intelligence Index | 44.3 |
| Artificial Analysis | Grok 4.6 (high) Coding Index | 76.8 |
| Artificial Analysis | Grok 4.6 (high) Agentic Index | 53.0 |
| Artificial Analysis | Grok 4.6 (high) GPQA Diamond | 94.9% |
| Artificial Analysis | Grok 4.6 (high) HLE | 42.9% |
| Artificial Analysis | Grok 4.6 (high) AA-LCR | 80.3% |
| Artificial Analysis | Grok 4.6 (high) GDPval-AA | 55.3% |
| Artificial Analysis | Grok 4.6 (high) CritPt | 17.1% |
| Artificial Analysis | Grok 4.6 (high) SciCode | 56.5% |
| Artificial Analysis | Grok 4.6 (high) AA-Omniscience Accuracy | 48.2% |
| Artificial Analysis | Grok 4.6 (high) AA-Omniscience Non-Hallucination Rate | 65.7% |
| Design Arena | Grok 4.6 Agents Arena Agenticgamedev Elo | 1210 |
| Design Arena | Grok 4.6 Agents Arena Androidnative Elo | 1302 |
| Design Arena | Grok 4.6 Agents Arena Full Stack Elo | 1274 |
| Design Arena | Grok 4.6 Agents Arena Htmlslides Elo | 1253 |
| Design Arena | Grok 4.6 Agents Arena Mobile Apps Elo | 1268 |
| Design Arena | Grok 4.6 Agents Arena Python-Pptxslides Elo | 1242 |
| Design Arena | Grok 4.6 Agents Arena Webapps Elo | 1255 |
| Design Arena | Grok 4.6 Models Arena 3D Elo | 1304 |
| Design Arena | Grok 4.6 Models Arena Asciiart Elo | 1297 |
| Design Arena | Grok 4.6 Models Arena Code Categories Elo | 1306 |
| Design Arena | Grok 4.6 Models Arena Data Visualization Elo | 1301 |
| Design Arena | Grok 4.6 Models Arena Game Development Elo | 1320 |
| Design Arena | Grok 4.6 Models Arena SVG Elo | 1261 |
| Design Arena | Grok 4.6 Models Arena UI Component Elo | 1304 |
| Design Arena | Grok 4.6 Models Arena Website Elo | 1301 |
Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick — Balanced (price + speed), Nitro (fastest), Floor (cheapest), or Exacto (highest tool-calling accuracy).
The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.
Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).
Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.
Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.
Token volume and request traffic to this model over time.
Drop-in code to call this model. OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. The only thing that changes between models is the model slug below.
| $2.00 | $6.00 | $0.50 | 1.08s | 50 tps | ||
| $2.00 | $6.00 | $0.50 | 0.64s | 51 tps | ||
Not used in Standard routing:Why these endpoints are not used | ||||||
 | $2.20 | $6.60 | $0.55 | 27.39s | 138 tps | |
Priority | $4.00 | $12.00 | $1.00 | 4.68s | 57 tps | |
Priority | $4.00 | $12.00 | $1.00 | -- | -- | |
Throughput
138tok/s
P50, best across providers
Latency
0.64s
P50, best provider
100.00%
98.83%
When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and customization options.
Grok 4.6 is a model from SpaceXAI with frontier performance on coding, knowledge work, and STEM. It is succeeded by Grok 4.7.
Grok 4.6 costs $2.00/M input tokens and $6.00/M output tokens, with separate rates for Cache Read at $0.50/M tokens and Web Search at $5.00/1K calls.
Grok 4.6 has a 500,000 token context window.
Yes. Grok 4.6 accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.
Grok 4.6 accepts text, images, and files such as PDFs as input and returns text.
Grok 4.6 is served by 2 providers on OpenRouter: SpaceXAI and Amazon Bedrock. Requests are routed to the best available provider, with automatic failover to the others, and you can pin or exclude providers with provider routing.
Grok 4.6 was released on August 12, 2026.