Compare startups
Pick two or three companies from the roster to see them side by side. The link in your address bar keeps the comparison, so you can send it to someone.
| Fact | InferenceInference.netRemove | InferenceMorphRemove | InferenceRelaceRemove |
|---|---|---|---|
| Sector | Inference | Inference | Inference |
| What they do | Hosts custom and open models, and traces and tunes the agents that call them. | Fast edit, search and context-compaction models for coding agents. | Fast small models and tools for coding agents, plus hosted open models. |
| Stage | Seed | Not on file | Series A |
| Open roles | 0 | 0 | 0 |
| Tracked since | October 2026 | October 2026 | October 2026 |
| Why we track it | Inference.net publishes per-token prices for dozens of hosted models, a free pay-as-you-go plan and an open-source tool that reads agent traces, so its cost and latency claims can be checked by hand. | Morph publishes a per-token price list, a free monthly allowance and its own explanation of how prefix caching behaves across an agent's long tool calls, so its speed and caching claims can be checked by hand. | Relace publishes a per-token price for every model it sells and explains in its own words how its context rewrite interacts with a provider's KV cache, so its speed and savings claims can be checked against a real agent trace. |
Following the roster
Run a startup?
The practice starts by doing the work: what we find goes to you first, and a free weekend on one named problem can follow.