Data as of October 23, 2026 · How the score is built
Release date Coming soonN/A contextBlock launch page
Dessa
Decision readingDessa is tracked, but not publicly ranked yet. Published speech or voice results from Dessa launch recordings appear below. They do not produce a weighted text-model score.
Voice benchmark evidence
Dessa launch recordings is an owner-defined voice evaluation. Its result stays separate from BenchLM's weighted text-model ranking.
Lineage
The sequence follows explicit supersedes links. Each score is estimated for that model; a relative can inform a sparse estimate but never sets a floor, so a newer release can score below an earlier one. Scores and prices remain blank when the corresponding public row or first-party rate is unavailable.
Base entry
Spec sheet
Each documented value carries its source. Missing fields stay visible as not sourced or not published, rather than disappearing from the page.
- API model ID
- Not publishedBlock model documentation
- Context window
- N/A
- Maximum output
- Not sourced yet
- Knowledge cutoff
- Not sourced yet
- Input modalities
- Not sourced yet
- Output modalities
- Not sourced yet
- Parameters
- Not sourced yet
- Availability
- Announced by Block with a waitlist and early access for a small group of AI safety researchers. Block says Dessa listens and speaks at the same time, runs faster than real time, and runs locally on a MacBook Pro M4 Max or better. No API model ID, weights, license, or release date is published yet.
- Cloud regions
- Not tracked yet
- Lifecycle
- Coming Soon
- API capabilities
- Tool calling, structured outputs, and batch support are not tracked yet
- Prompt caching
- Not documented in the pricing recordBlock model documentation
- Self-host
- Weight availability is pending
- Rate limits
- Not tracked yet
How to read this profile
The visual layer above carries the decisions. These notes preserve the model, ranking, coverage, and family context behind the numbers.
We track Dessa, but no weighted text-model benchmark result is published on the site yet. This page shows the metadata and separate protocol evidence we can verify now; a BenchLM score will appear only if compatible public evaluations land.
Dessa has no verified publication yet for weight access. Its N/A context window is documented separately from that pending weight status. The profile records its reasoning mode as non-reasoning.
Announced by Block with a waitlist and early access for a small group of AI safety researchers. Block says Dessa listens and speaks at the same time, runs faster than real time, and runs locally on a MacBook Pro M4 Max or better. No API model ID, weights, license, or release date is published yet.
Dessa is a full-duplex speech-to-speech conversation model from a small research lab at Block, trained from scratch on human speech only, with no text and no pretrained models. Block has announced it with a waitlist, so it is listed as coming soon and stays separate from weighted text-model scores.
Speech or voice evidence appears above; text-model benchmark coverage remains empty.
Last updated October 23, 2026. Runtime fields remain blank until a sourced snapshot exists.
Questions
How does Dessa perform overall in AI benchmarks?
Dessa has published speech or voice results in Dessa launch recordings. Those results remain separate from the weighted text-model ranking, so this profile does not assign a BenchLM score or overall rank. Use the linked evaluation's metric and run conditions to compare speech systems.
Does Dessa have full benchmark coverage on BenchLM?
No. Dessa's speech or voice results appear in Dessa launch recordings, but it has no published text-model benchmark rows. The voice evaluation does not fill the text-model coverage slots or produce an overall rank. Missing text coverage is not treated as a zero score.
What is the context window size of Dessa?
Dessa has a reported context window of N/A in the exact-model catalog record. The value stays visible, but the profile marks its source link as unavailable instead of presenting it as directly documented. Maximum output length remains separate because providers often publish a different limit.