# General Compute vs Boundless

> Both sell open models and reserved capacity under contract. General Compute moves decode onto ASICs, and Boundless tunes the serving stack per customer.

Canonical: https://www.subconscious.dev/compare/general-compute-vs-boundless · By The Subconscious Team · Updated October 29, 2026

## How they compare

General Compute runs prefill on NVIDIA GPUs and decode on ASICs from Cerebras, SambaNova and others, and in its own benchmark measured over 1,000 tokens per second on GPT-OSS 120B. Its OpenAI compatible API in us-west-2 serves four older open models, from DeepSeek V3.2 with 32K context to MiniMax M2.7 with 192K, at $0.21–$0.28 in per million tokens, and hosts LoRA or fine-tuned checkpoints. Boundless lists 22 models, including GLM-5.3 and Kimi K3 with 1M context and resold GPT and Claude models, behind an API that works with the OpenAI and Anthropic SDKs. It runs requests in the US or Europe and offers no fine-tuning. Vercel AI Gateway measures it at 342 tokens per second on DeepSeek V4.1 Flash 1017.

Pick General Compute when decode speed matters most, or to host a fine-tuned checkpoint. Pick Boundless for newer models with long context, an Anthropic compatible endpoint or European processing. Both sell reserved capacity under contract. Boundless lets you test it free for two weeks on real traffic, and most of General Compute's dedicated capacity goes live in Q1 2027.

## What each one does

### General Compute

General Compute is a San Francisco inference cloud built on alternative chips. It splits each request across hardware, with NVIDIA GPUs handling prefill and ASICs from Cerebras, SambaNova, Positron AI and d-Matrix handling decode. In its own open benchmark on GPT-OSS 120B it measured 1,025 to 1,234 output tokens per second against 100 to 121 on an unnamed GPU cloud, and up to 16.2x faster end to end. Founders Finn Puklowski and Jason Goodison raised a $15M seed led by FUSE, an up to $400M debt facility from Upper90 and, on October 27, 2026, a $100M Series A co-led by Blackbird and Abstract.

### Boundless

Boundless launched in September 2026 as an inference partner for AI-native startups. It sells two things. The Model API serves pre-tuned models on shared capacity, billed per token from prepaid credits with a budget per API key. Tailored inference reserves capacity for one customer, and Boundless engineers pick the model on the customer’s own evals, tune precision, batching, caching, speculative decoding and parallelism, then run the deployment against uptime, throughput and latency targets written into the contract. A free two-week evaluation tests those targets on real traffic against the customer’s current provider first.

## Which is best, and when

### Choose General Compute for

- Decode speed on ASICs
- Hosting LoRA and fine-tuned checkpoints
- Tokens-per-second SLAs on dedicated racks

### Choose Boundless for

- Newer open models with 1M context
- OpenAI and Anthropic SDK support
- Processing in the US or Europe

## At a glance

| Attribute | General Compute | Boundless |
|---|---|---|
| Model access | Open weights | Open weights, plus closed GPT and Claude |
| Flagship models | MiniMax M2.7, DeepSeek V3.2, GPT-OSS 120B | GLM-5.3, DeepSeek V4.1 Flash, Kimi K3 |
| Speed | 1,000+ tok/s on GPT-OSS 120B (vendor) | 342 tok/s on DeepSeek V4.1 Flash 1017 (Vercel) |
| Price | $0.21–$0.28 in, $0.38–$1.20 out per 1M | $0.04–$2.30 in, $0.21–$11.40 out per 1M (open) |
| Customization | Custom LoRA and fine-tuned checkpoints | Serving-stack tuning, no fine-tuning |
| Deployment | Serverless API, dedicated racks, bare metal | Serverless API, reserved tailored capacity |
| Long context | 192K (MiniMax M2.7) | 1M context on most models |

## FAQ

### What is the difference between General Compute and Boundless?

Both sell open models and reserved capacity under contract. General Compute moves decode onto ASICs, and Boundless tunes the serving stack per customer.

### When should I choose General Compute over Boundless?

Decode speed on ASICs; Hosting LoRA and fine-tuned checkpoints; Tokens-per-second SLAs on dedicated racks.

### When should I choose Boundless over General Compute?

Newer open models with 1M context; OpenAI and Anthropic SDK support; Processing in the US or Europe.

### Is General Compute or Boundless cheaper?

General Compute: $0.21–$0.28 in, $0.38–$1.20 out per 1M. Boundless: $0.04–$2.30 in, $0.21–$11.40 out per 1M (open). The cheaper choice depends on the model and workload.

### Which has more context, General Compute or Boundless?

General Compute: 192K (MiniMax M2.7). Boundless: 1M context on most models.

## Running long-horizon agents?

If your agents run past 200K tokens, compare both against Subconscious: [Subconscious vs General Compute](https://www.subconscious.dev/compare/subconscious-vs-general-compute.md), [Subconscious vs Boundless](https://www.subconscious.dev/compare/subconscious-vs-boundless.md).

Full profiles: [General Compute](https://www.subconscious.dev/providers/general-compute.md), [Boundless](https://www.subconscious.dev/providers/boundless.md).