# Cohere vs Inception

> Cohere builds enterprise models for RAG and agents that can run in your own VPC. Inception builds diffusion LLMs aimed at speed.

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

## How they compare

Cohere's Command A+ is a 218B mixture-of-experts model under Apache 2.0 with 128K context, and Command A lists at $2.50 in and $10 out with 256K context. Embed 5 and Rerank 4 make a mature retrieval stack, and private VPC or on-prem deployment with fine-tuning is a core product. 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.

Cohere is the pick for regulated enterprises that need retrieval and agents inside their own network, or open weights to host. Inception is the pick for latency-bound steps like voice and code edits. They also combine: Cohere's Embed and Rerank for retrieval, Mercury to write the answer fast.

## What each one does

### Cohere

Cohere is a Toronto-based lab that sells models and platforms to banks, governments and large enterprises rather than consumers. Its generative line is the Command family. Command A+, released May 20, 2026, is a 218B-parameter mixture-of-experts model with 25B active, published under Apache 2.0 with a 128K context window, and it combines reasoning, vision, translation and tool use in one set of weights. Command A has a 256K window and lists at $2.50 in and $10 out per million tokens, while Command R7B costs $0.0375 in. June 2026 added North Mini Code, a 30B Apache 2.0 coding model, and the lineup also includes Aya multilingual models and Transcribe for speech.

### 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 Cohere for

- RAG with Embed and Rerank
- Private VPC or on-prem deployment
- Open Command A+ weights under Apache 2.0

### Choose Inception for

- Fast answer generation for chat and voice
- Code autocomplete and edits
- Low per-token prices with tool use

## At a glance

| Attribute | Cohere | Inception |
|---|---|---|
| Model access | Closed, plus open Command A+ | Closed weights (Mercury dLLMs) |
| Flagship models | Command A+, Command A, Embed 5, Rerank 4 | Mercury 2.5, Mercury 2 |
| Speed | 375 tok/s on Command A+ W4A4, per Cohere | 1,100+ tok/s (vendor); ~640 measured |
| Price | $0.0375–$2.50 in, $0.15–$10 out per 1M | $0.20 in, $0.75 out (Mercury 2.5) |
| Customization | Enterprise fine-tuning, incl. private | Fine-tuning on request |
| Deployment | API, Bedrock, Azure, OCI, VPC, on-prem | Serverless API, Bedrock, Azure, private |
| Long context | 256K on Command A; 128K on A+ | 260K |

## FAQ

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

Cohere builds enterprise models for RAG and agents that can run in your own VPC. Inception builds diffusion LLMs aimed at speed.

### When should I choose Cohere over Inception?

RAG with Embed and Rerank; Private VPC or on-prem deployment; Open Command A+ weights under Apache 2.0.

### When should I choose Inception over Cohere?

Fast answer generation for chat and voice; Code autocomplete and edits; Low per-token prices with tool use.

### Is Cohere or Inception cheaper?

Cohere: $0.0375–$2.50 in, $0.15–$10 out per 1M. Inception: $0.20 in, $0.75 out (Mercury 2.5). The cheaper choice depends on the model and workload.

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

Cohere: 256K on Command A; 128K on A+. Inception: 260K.

## Running long-horizon agents?

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

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