Exa vs Letta
Exa scores higher on the AgentReady, 92/100 against 72/100. They differ on 9 of the 41 signals. Which ones decides whether an agent can adopt them without a person watching.
What each one is
Exa. Exa is a custom search engine built for AIs, offering an API to search the largest index of public web and private information.
Letta. Letta is an AI research lab in San Francisco building machines that learn.
Where Exa is ahead
Exa passes structured api reference, openapi / spec quality, request examples provided, response examples provided and errors and status codes documented, and Letta does not. That is understand, whether an agent can read the docs and work out how the API behaves before calling it.
It also holds adopt: programmatic credential creation. Letta misses it.
And on operate, machine-readable errors and rate-limit behavior predictable. Letta misses those.
Where Letta is ahead
Letta passes agent-compatible signup flow, and Exa does not. That is adopt, whether an agent can get a key and make its first successful call without a human in the loop.
What neither does
Both fail no mandatory sales call, retry behavior documented, idempotency support. If your agent needs any of those, you will be building it yourself either way.
Score, pillar by pillar
The AgentReady splits into four pillars, scored separately, because a product can be easy to find and still impossible to adopt.
Understand is whether an agent can read the docs and work out how the API behaves before calling it. Exa leads 100 to 38. Exa misses nothing; Letta misses structured api reference, openapi / spec quality, request examples provided, response examples provided, errors and status codes documented.
Adopt. Both sit at 81/100 here. Exa misses no mandatory sales call, agent-compatible signup flow; Letta misses no mandatory sales call, programmatic credential creation.
Operate. Exa leads 88 to 67. Exa misses retry behavior documented, idempotency support; Letta misses machine-readable errors, retry behavior documented, idempotency support, rate-limit behavior predictable.
Pricing
Exa starts at $0 (free tier with credits) and has a free tier. Letta starts at $0/mo and has a free tier.
| Exa plans | Letta plans |
|---|---|
| - | Free $0 /month |
| - | Pro $20 /month |
| - | API Plan $20 /month |
| - | Teams Pro $20 /seat/month |
| - | Enterprise Custom |
Signal by signal
| Signal | Exa | Letta |
|---|---|---|
| AgentReady | 92 | 72 |
| Discovery | 100 | 100 |
| Understanding | 100 | 38 |
| Adoption | 81 | 81 |
| Operability | 88 | 67 |
| Public API | Yes | Yes |
| MCP server | Yes | Yes |
| OpenAPI spec | Yes | Unknown |
| CLI | Yes | Yes |
| llms.txt | Yes | Yes |
| Self-serve signup | Yes | Yes |
| Free tier | Yes | Yes |
Which to pick
Exa clears more of the signals an agent needs, so it is the safer default for unattended use. Full profiles: Exa and Letta. Alternatives to each: Exa, Letta.
An agent can fetch this as data: POST /v1/compare {"slugs": ["exa", "letta"]}