vs
DeepInfra vs Inception
DeepInfra is the price floor for 150+ open models. Inception sells closed diffusion LLMs priced low and built for speed.
By The Subconscious Team · Updated
DeepInfra vs Inception: key differences
DeepInfra serves Llama 3.1 8B at $0.02 per million and DeepSeek V4 Flash at $0.09 in and $0.18 out, with no minimums or contracts. Heavy quantization helps its prices but can cut quality and context, on some endpoints. Inception's Mercury 2.5 costs $0.20 in and $0.75 out with 260K context and tool use. Inception reports over 1,100 tokens per second, and Artificial Analysis measured about 640.
DeepInfra wins on price and model choice for bulk jobs where speed matters little. Inception costs more per token but returns output much faster, which matters for user-facing loops like code edits, chat and voice. DeepInfra has no managed fine-tuning, while Inception offers it on request.
What DeepInfra and Inception do
DeepInfra
DeepInfra is the price floor for open-model inference. Developers treat it as the reference point for what a token should cost, with small models like Llama 3.1 8B at $0.02 per million and DeepSeek V4 Flash at $0.09 in and $0.18 out. The catalog covers 150+ open models across text, image and speech behind a fully OpenAI-compatible API. There are no minimums, setup fees or contracts on the shared API.
Example models: DeepSeek V4 Flash, Llama 3.1 8B
Full DeepInfra profile
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.
Example models: Mercury 2.5, Mercury 2
Full Inception profileShould you choose DeepInfra or Inception?
DeepInfra vs Inception at a glance
| Attribute | DeepInfra |  Inception |
|---|---|---|
| Model access | Open weights | Closed weights (Mercury dLLMs) |
| Flagship models | DeepSeek V4 Flash, Llama 3.1 8B | Mercury 2.5, Mercury 2 |
| Speed | Unknown | 1,100+ tok/s (vendor); ~640 measured |
| Price | From $0.02 per 1M | $0.20 in, $0.75 out (Mercury 2.5) |
| Customization | No managed fine-tuning | Fine-tuning on request |
| Deployment | Shared API, no contracts | Serverless API, Bedrock, Azure, private |
| Long context | 1M on DeepSeek V4 Pro | 260K |
Frequently asked questions
What is the difference between DeepInfra and Inception?
DeepInfra is the price floor for 150+ open models. Inception sells closed diffusion LLMs priced low and built for speed.
When should I choose DeepInfra over Inception?
Lowest per-token cost on open models; Bulk extraction, tagging and synthetic data; Broad catalog with fast new releases.
When should I choose Inception over DeepInfra?
User-facing loops where speed matters; Code edits and voice agents; Reasoning with tool use and 260K context.
Is DeepInfra or Inception cheaper?
DeepInfra: From $0.02 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, DeepInfra or Inception?
DeepInfra: 1M on DeepSeek V4 Pro. Inception: 260K.
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.