Flint
Recursive's 40B coding model — only about 2.5B parameters active at a time — on one desktop-sized AI computer.
Setup and speed
Full results
- —Each task is run three times; the score counts all 1,650 attempts, not 550 tasks.
- —Built for coding agents, yet it posts a perfect 50/50 on classification and routing and 97–98 on inbox triage and document extraction.
- —The weak spots are the multi-step number work: Spreadsheet analysis 82 and Formulas & SQL 83, both dropping to 56.7 on hard tasks. On spreadsheets it most often ran out of its twelve tool steps while still working (37 attempts), rather than giving a wrong answer.
- —461 of 550 tasks were correct on all three passes. Nothing was cut off by the output limit.
With thinking on, the first token arrives at 0.21s but the visible answer does not start until roughly 16s while the model reasons. The 38.6s typical task time is that reasoning plus the answer.
Files it produced
Four real documents from one fact sheet, produced once. Passed is machine-checked against the fact sheet and the brief; quality is a separate 1–5 score from a published rubric and never changes whether a file passed. How files are graded →
An 8–12 page annual report with six required sections: shareholder letter, financial review, category performance, customer and geographic performance, risk factors, and outlook.
Download DOCXA two-sheet workbook: a Summary comparing FY2025 to FY2024, and a Monthly sheet with all twelve months.
Download XLSXA ten-slide investor presentation following a fixed outline, with the brand palette applied and at least one real chart object.
Download PPTXAn X post of 280 characters or fewer and a LinkedIn post, both announcing the results, text only.
Download MDEvery artifact below is the file the model actually produced, published unedited so you can open it yourself. These are AI-generated documents about a fictional company — read them as samples of the model's work, not as advice or a real annual report. The workbook's formula cells hold real formulas — a viewer that does not recalculate may show them blank until Excel opens the file.