# Baseten vs Inception

> Baseten serves about a dozen open models and any custom model you package, with the lowest measured time to first token. Inception serves its own closed diffusion LLMs built for fast output.

Canonical: https://www.subconscious.dev/compare/baseten-vs-inception · By The Subconscious Team · Updated October 11, 2026

## How they compare

Baseten's Model APIs cover GLM 5.3, DeepSeek V4, Kimi K3, gpt-oss 120B and others over OpenAI and Anthropic-compatible endpoints, with a 0.49 second time to first token on Artificial Analysis. Dedicated deployments run any model through Truss, billed per GPU minute with a 99.99% uptime SLA. Inception's Mercury 2.5 is a closed reasoning model with 260K context, tool use and JSON mode at $0.20 in and $0.75 out per million. Inception reports over 1,100 tokens per second, and Artificial Analysis measured about 640.

Baseten wins on choice: open models, your own fine-tunes, self-hosting and HIPAA options. Inception's draw is fast output at low prices, but only on Mercury, whose quality trails large open models on hard reasoning. A fast first token suits short chat turns, and fast output suits long generations like code edits.

## What each one does

### 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.

### Inception

Inception builds diffusion large language models (dLLMs). Every other LLM on this list generates text one token at a time; Inception's Mercury models start from a rough draft of a whole block and refine it over a few passes, producing many tokens in parallel. That ties speed less to output length and lets the models run fast on ordinary NVIDIA GPUs rather than custom chips. The company was founded by professors from Stanford, UCLA and Cornell behind core work on diffusion, Flash Attention and DPO, and is led by Stanford's Stefano Ermon. It raised a $50M seed in November 2025 led by Menlo Ventures, with Mayfield, M12, NVentures, Snowflake and Databricks, plus Andrew Ng and Andrej Karpathy as angels.

## Which is best, and when

### Choose Baseten for

- Low time to first token on open models
- Serving private fine-tunes and custom models
- Regulated deployments with HIPAA and self-host

### Choose Inception for

- Fast long outputs like code edits
- Voice agents and real-time chat
- Low-cost reasoning with tool use

## At a glance

| Attribute | Baseten | Inception |
|---|---|---|
| Model access | Open weights, about a dozen curated | Closed weights (Mercury dLLMs) |
| Flagship models | GLM 5.3, DeepSeek V4.1 Flash, Kimi K3, gpt-oss 120B | Mercury 2.5, Mercury 2 |
| Speed | 0.49s TTFT, lowest measured | 1,100+ tok/s (vendor); ~640 measured |
| Price | H100 about $6.50/hr dedicated | $0.20 in, $0.75 out (Mercury 2.5) |
| Customization | Deploy any model with Truss | Fine-tuning on request |
| Deployment | Model APIs, dedicated, self-host | Serverless API, Bedrock, Azure, private |
| Long context | Varies by model | 260K |

## FAQ

### What is the difference between Baseten and Inception?

Baseten serves about a dozen open models and any custom model you package, with the lowest measured time to first token. Inception serves its own closed diffusion LLMs built for fast output.

### When should I choose Baseten over Inception?

Low time to first token on open models; Serving private fine-tunes and custom models; Regulated deployments with HIPAA and self-host.

### When should I choose Inception over Baseten?

Fast long outputs like code edits; Voice agents and real-time chat; Low-cost reasoning with tool use.

### Is Baseten or Inception cheaper?

Baseten: H100 about $6.50/hr dedicated. Inception: $0.20 in, $0.75 out (Mercury 2.5). The cheaper choice depends on the model and workload.

### Which has more context, Baseten or Inception?

Baseten: Varies by model. Inception: 260K.

## Running long-horizon agents?

If your agents run past 200K tokens, compare both against Subconscious: [Subconscious vs Baseten](https://www.subconscious.dev/compare/subconscious-vs-baseten.md), [Subconscious vs Inception](https://www.subconscious.dev/compare/subconscious-vs-inception.md).

Full profiles: [Baseten](https://www.subconscious.dev/providers/baseten.md), [Inception](https://www.subconscious.dev/providers/inception.md).