The AI arena is free today

Open Superagent
API Provider12 active models1 organization

xAI: API pricing, speed & models

xAI hosts 12 active AI models, with input pricing from $0.20 per 1M tokens, with median throughput of 37 characters/sec, and P95 time to first token of 1.65s, with 67.8% success rate over 7 days. Compare xAI's API speed, pricing, and reliability against other inference providers.

Image generationAudioVideoFirst-party only
Median throughput37 c/s
Median TTFT1.19s
P95 TTFT1.65s
Success rate (7d)67.8%
From$0.20 /M tok

Catalog

Type
Price
38 models
Model
Grok 4.6cache pricing
Grok 4.6cache pricing
Grok 4.5cache pricing · 80 QPS
Grok 4.5cache pricing · 80 QPS
Grok-3100 QPS
Grok-3100 QPS
At a glance

xAIpricing, performance & catalog

The citable facts about xAI's 12 models — sourced from provider APIs and refreshed continuously.

Lowest price
Grok-4 Fast Reasoning at $0.200 per 1M input tokens
Highest median throughput
Grok 4.6 at 54 chars/s
Lowest median TTFT
Grok 4.3 at 1.15s
Largest context
Grok-4 Fast Reasoning at 2.0M tokens
Catalog
12 active models from 1 organization

FAQ

Common questions about xAI.

What is xAI?

xAI is an API provider that hosts large language models. Active models: 12; From (input): $0.20 / 1M tok; Median throughput: 37 c/s; Median TTFT: 1.19s; Success rate (7d): 67.8%.

How many models does xAI offer?

xAI currently serves 12 active models out of 24 historical offerings on LLM Stats.

What is xAI's API pricing?

xAI input pricing starts from $0.20 per 1M tokens, with the most expensive offering at $3 per 1M tokens. See the Pricing tab above for the full per-model breakdown.

How fast is xAI?

xAI delivers a median throughput of 37 characters per second and P95 time to first token of 1.65s across its model catalog. See the Models tab for per-model throughput and TTFT breakdowns.

Is xAI reliable?

xAI has a 67.8% success rate across 105 API calls in the last 7 days, with a 32.2% error rate.

Does xAI support function calling?

Yes. 14 of 12 models on xAI support function calling (tool use). The Capabilities tab lists which specific models accept tool definitions.

Does xAI support JSON mode and structured output?

Yes. 14 of 12 models on xAI support structured output (JSON mode / schema-constrained generation). The Capabilities tab shows which specific models accept response_format or json_schema parameters.

Does xAI offer batch inference?

Yes. 4 of 12 models on xAI support batch inference for cheaper, asynchronous workloads.

Does xAI support multimodal models?

Yes. xAI's catalog includes 15 vision-capable, 6 image generation, 6 audio, and 12 video models. See the Models and Capabilities tabs for the full per-model breakdown.

Whose models does xAI host?

xAI hosts models from xAI. See the Models tab for the full catalog grouped by creator.

How do I start using xAI?

Sign up at https://docs.x.ai to get an API key, then call xAI's API directly from your application. Most clients work out of the box by pointing the OpenAI SDK at xAI's base URL with your key. Use the Models and Pricing tabs above to pick the right model for your latency, cost, and context-window requirements.