# xAI vs Crusoe

> xAI sells closed Grok models with live X data and cheap output tokens. Crusoe serves open models on its own hardware, with fine-tuning and raw GPUs.

Canonical: https://www.subconscious.dev/compare/xai-vs-crusoe · By The Subconscious Team · Updated September 30, 2026

## How they compare

xAI's Grok 4.6 costs $2 in and $6 out per million tokens with a 500K context, and Grok 4.20 and 4.3 keep a 1M window at $1.25 in and $2.50 out. Once a prompt passes 200K tokens, the whole request bills at double. Crusoe serves open models only, including DeepSeek, GLM 5.3, Kimi K2.6, Gemma, gpt-oss and Nemotron, from $0.05 in and $0.20 out up to $1.74 in and $4.40 out per million, with cached input well below list. For agent contexts that reread the same prefix, Crusoe's cluster-wide KV cache is the relevant feature. Grok 4.6 runs around 54 tokens per second. Crusoe's speed claim, up to 9.9x faster time to first token versus vLLM, is its own benchmark on prefix-heavy work.

xAI's distinct advantage is data. Server-side Web Search and X Search tools pull in current posts straight from X, which suits news, market and social-sentiment agents, and xAI ships separate APIs for image, video and audio plus a multi-agent Grok 4.20 variant. It offers no customization and runs only on its first-party API. Crusoe offers serverless LoRA fine-tuning, dedicated endpoints per GPU-hour, tailored deployments with SLAs and raw GPU clusters on Kubernetes or Slurm. Its downside is a small serverless catalog. Pick xAI for real-time social context and cheap closed-model output. Pick Crusoe for open weights you can fine-tune, dedicated capacity, and long agent runs that should not bill at double past 200K.

## What each one does

### xAI

xAI sells the Grok models through its own API. Grok 4.6 is the current flagship and xAI tells developers to use it for everything outside audio, image and video, code included. It has a 500K context window and costs $2 in and $6 out per million tokens under 200K prompt tokens. Older Grok 4.20 and 4.3 models keep a 1M window at $1.25 in and $2.50 out, which is aggressive for that capability class.

### Crusoe

Crusoe started in 2018 turning wasted natural gas into power for computing and has since become a vertically integrated AI infrastructure company: it sources energy, builds data centers and rents GPUs through Crusoe Cloud. It designed and built the Abilene, Texas campus behind the OpenAI and Oracle Stargate project, planned at 1.2 GW, and in March 2026 announced an adjacent 900 MW campus for Microsoft. On September 17, 2026 it closed the first part of a $3.9B Series F at a $30.9B post-money valuation, and it reports over 6 GW of contracted capacity. Crusoe Cloud lists GB200 NVL72, B200 and AMD MI355X by quote, with H100 at $3.90 and H200 at $4.29 per GPU-hour on demand.

## Which is best, and when

### Choose xAI for

- Social listening with native X Search
- News and market agents that need fresh data
- Image, video and audio from the same vendor

### Choose Crusoe for

- Open weights with LoRA fine-tuning
- Long agent contexts without Grok's 2x rate past 200K
- Dedicated endpoints and raw GPU clusters

## At a glance

| Attribute | xAI | Crusoe |
|---|---|---|
| Model access | Closed | Open weights |
| Flagship models | Grok 4.6, Grok 4.20, grok-build | DeepSeek V4, GLM 5.3, Kimi K2.6, Nemotron 3 |
| Speed | ~54 tok/s on Grok 4.6 | Up to 9.9x faster TTFT vs vLLM (vendor claim) |
| Price | $2 in, $6 out (Grok 4.6); 2x past 200K | $0.05–$1.74 in, $0.20–$4.40 out per 1M |
| Customization | - | Serverless LoRA fine-tuning |
| Deployment | First-party API | Serverless, self-serve and tailored dedicated, raw GPUs |
| Long context | 500K (4.6), 1M (4.20, 4.3) | Varies by model; cluster-wide KV cache |

## FAQ

### What is the difference between xAI and Crusoe?

xAI sells closed Grok models with live X data and cheap output tokens. Crusoe serves open models on its own hardware, with fine-tuning and raw GPUs.

### When should I choose xAI over Crusoe?

Social listening with native X Search; News and market agents that need fresh data; Image, video and audio from the same vendor.

### When should I choose Crusoe over xAI?

Open weights with LoRA fine-tuning; Long agent contexts without Grok's 2x rate past 200K; Dedicated endpoints and raw GPU clusters.

### Is xAI or Crusoe cheaper?

xAI: $2 in, $6 out (Grok 4.6); 2x past 200K. Crusoe: $0.05–$1.74 in, $0.20–$4.40 out per 1M. The cheaper choice depends on the model and workload.

### Which has more context, xAI or Crusoe?

xAI: 500K (4.6), 1M (4.20, 4.3). Crusoe: Varies by model; cluster-wide KV cache.

## Running long-horizon agents?

If your agents run past 200K tokens, compare both against Subconscious: [Subconscious vs xAI](https://www.subconscious.dev/compare/subconscious-vs-xai.md), [Subconscious vs Crusoe](https://www.subconscious.dev/compare/subconscious-vs-crusoe.md).

Full profiles: [xAI](https://www.subconscious.dev/providers/xai.md), [Crusoe](https://www.subconscious.dev/providers/crusoe.md).