Best AI Model for Finance (November 2026) — Live SOTA Ranking
As of Nov 3, 2026, no model holds the Finance crown on isitsota. Muse Spark 1.2 ranks #1 on the board, but a crown needs a confirmed lead. The ranking is a weighted score over 4 benchmarks, recomputed automatically (methodology v1.0).
Last updated Nov 3, 2026 · Full Finance board: crown history, benchmarks and methodology
Top 10 ranked
| Rank | Model | Lab | Status | Score | Open weights | Price |
|---|---|---|---|---|---|---|
| 1 | Muse Spark 1.2 | Meta | Near-SOTA · Finance | 100.0 | No | $2 / 1M tokens (blended) |
| 2 | Muse Spark 1.1 | Meta | Near-SOTA · Finance | 99.1 | No | $2 / 1M tokens (blended) |
| 3 | Muse Spark | Meta | Near-SOTA · Finance | 98.1 | No | — |
| 4 | Claude Sonnet 4.6 | Anthropic | Near-SOTA · Finance | 97.2 | No | $6 / 1M tokens (blended) |
| 5 | Claude Fable 5 | Anthropic | Near-SOTA · Finance | 96.3 | No | $20 / 1M tokens (blended) |
| 1 | Claude Opus 5.5 | Anthropic | #1 in Finance | — | No | $8 / 1M tokens (blended) |
| 2 | Gemini 4 Argon | Google DeepMind | #2 in Finance | — | No | $4 / 1M tokens (blended) |
| 3 | Muse Wonder Full | Meta | #3 in Finance | — | No | — |
| 6 | GPT-5.6 Luna | OpenAI | #6 in Finance | 94.4 | No | $0.45 / 1M tokens (blended) |
| 6 | GPT-5.6 Terra | OpenAI | #6 in Finance | 94.4 | No | $4.5 / 1M tokens (blended) |
Score is the weighted category score (the top model scores 100). Prices are the latest observed list prices.
Budget, open-weights and local picks
- Best on a budget: GPT-5.6 Luna. Cheapest of the 8 priced models in the top 10: $0.45 per 1M tokens, ranked #6.
- Best open-weights: Kimi K3. #1 on the open-weights board (license: Proprietary). See the board.
- Best to run locally (≤32B): Qwen3.8 27B. #1 among models at or under 32B, 27B parameters. See the board.
Ranks
Last crown changes
Crown changes for this board are still empty as of Nov 3, 2026.
Which benchmarks count
- Finance Agent — primary, weight 1
- Excel Modeling Benchmark (EMB) — secondary, weight 0.35
- TaxEval — secondary, weight 0.35
- Tax Agent Bench — secondary, weight 0.35
How we decide
Finance is ranked by a weighted score over 4 benchmarks: Finance Agent (primary, weight 1), Excel Modeling Benchmark (EMB) (secondary, weight 0.35), TaxEval (secondary, weight 0.35), and Tax Agent Bench (secondary, weight 0.35).
Only validated results count. A new #1 takes the crown once its lead is confirmed by a second signal; until then the page says the crown is pending.
The verdict engine re-checks its sources every hour and records each crown change with its date.
Popular comparisons
Who is SOTA in Finance?
As of Nov 3, 2026, Finance is uncrowned on isitsota (methodology v1.0).
Since when is the Finance crown current?
As of Nov 3, 2026, Finance has an open crown.
What is the best open-weights model for Finance?
As of Nov 3, 2026, Kimi K3: #1 on the open-weights board (license: Proprietary).
What is the best Finance model I can run locally?
As of Nov 3, 2026, Qwen3.8 27B: #1 among models at or under 32B, 27B parameters.
What is the cheapest top-10 model for Finance?
As of Nov 3, 2026, GPT-5.6 Luna. Cheapest of the 8 priced models in the top 10: $0.45 per 1M tokens, ranked #6.
Which benchmarks decide Finance?
Finance is ranked by a weighted score over 4 benchmarks: Finance Agent (primary, weight 1), Excel Modeling Benchmark (EMB) (secondary, weight 0.35), TaxEval (secondary, weight 0.35), and Tax Agent Bench (secondary, weight 0.35).
How often is this ranking updated?
Every hour, automatically. The board last changed on Nov 3, 2026.