Baseten vs General Compute

Baseten leads on time to first token with OpenAI and Anthropic-compatible APIs. General Compute targets decode speed on ASICs.

By The Subconscious Team · Updated October 27, 2026

Baseten vs General Compute: key differences

Baseten posted the lowest time to first token on the Artificial Analysis board in August 2026, 0.49 seconds, and serves about a dozen open models including GLM 5.3, Kimi K3 and gpt-oss 120B over endpoints that speak both OpenAI and Anthropic formats. Dedicated deployments run any Truss-packaged model, with an H100 at about $6.50 an hour. General Compute pairs GPU prefill with ASIC decode and reports over 1,000 tokens per second on GPT-OSS 120B in its own benchmark. Its API is OpenAI-only and serves four models at $0.21 to $0.28 in per million tokens.

Baseten is the pick when first-token latency, Anthropic-format compatibility or custom non-LLM models matter. General Compute fits workloads where output length sets the wait, since its pitch is decode throughput. Both host custom fine-tunes. Baseten runs them on GPUs you choose, and General Compute runs them behind a private model ID on its mixed hardware.

What Baseten and General Compute do

Baseten

Baseten runs two products. Model APIs serve a curated set of about a dozen open models, including GLM 5.3, DeepSeek V4.1 Flash, DeepSeek V4 Pro 0813, Kimi K3 and gpt-oss 120B, over endpoints that speak both the OpenAI Chat Completions shape and the Anthropic Messages shape. That dual compatibility means an existing OpenAI or Claude SDK, or a coding agent, points at Baseten with a base URL change. Dedicated deployments take any model you package with the open-source Truss CLI and bill per GPU minute, with an H100 at about $6.50 an hour.

Example models: GLM 5.3, gpt-oss 120B

Full Baseten profile

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.

Example models: MiniMax M2.7, DeepSeek V3.2, GPT-OSS 120B

Full General Compute profile

Should you choose Baseten or General Compute?

Baseten

Choose Baseten for

Lowest time to first token
Claude SDKs and coding agents via Anthropic format
Custom speech and embedding models

General Compute

Choose General Compute for

High decode throughput
Cheap GPT-OSS 120B
Dedicated capacity with throughput SLAs

Baseten vs General Compute at a glance

Attribute	Baseten	General Compute
Model access	Open weights, about a dozen curated	Open weights
Flagship models	GLM 5.3, DeepSeek V4.1 Flash, Kimi K3, gpt-oss 120B	MiniMax M2.7, DeepSeek V3.2, GPT-OSS 120B
Speed	0.49s TTFT, lowest measured	1,000+ tok/s on GPT-OSS 120B (vendor)
Price	H100 about $6.50/hr dedicated	$0.21–$0.28 in, $0.38–$1.20 out per 1M
Customization	Deploy any model with Truss	Custom LoRA and fine-tuned checkpoints
Deployment	Model APIs, dedicated, self-host	Serverless API, dedicated racks, bare metal
Long context	Varies by model	192K (MiniMax M2.7)

Frequently asked questions

What is the difference between Baseten and General Compute?

Baseten leads on time to first token with OpenAI and Anthropic-compatible APIs. General Compute targets decode speed on ASICs.

When should I choose Baseten over General Compute?

Lowest time to first token; Claude SDKs and coding agents via Anthropic format; Custom speech and embedding models.

When should I choose General Compute over Baseten?

High decode throughput; Cheap GPT-OSS 120B; Dedicated capacity with throughput SLAs.

Is Baseten or General Compute cheaper?

Baseten: H100 about $6.50/hr dedicated. General Compute: $0.21–$0.28 in, $0.38–$1.20 out per 1M. The cheaper choice depends on the model and workload.

Which has more context, Baseten or General Compute?

Baseten: Varies by model. General Compute: 192K (MiniMax M2.7).

Related comparisons

Running long-horizon agents?

If your agents run past 200K tokens, compare both against Subconscious. Our inference stack treats a long-horizon trace as the primary workload, so speed, cost, and accuracy hold up deep into the trace.

Subconscious vs Baseten Subconscious vs General Compute