1stTop pick
Researched 15d agoGemini 2.5 Flash
- BFCL
- 56.24%
- Output (from)
- $2.50 / 1M
Best LLMs for JSON / Tool use (2026)
As of 2026-10-29, LLM Reference's top pick for function calling and tool use is Gemini 2.5 Flash by Google DeepMind (BFCL 56.24%, $2.50/1M output tokens), chosen from models that meet the podium eligibility rules. Second pick: GPT-4.1 by OpenAI.
Last refreshed 2026-10-29. Next refresh: weekly.
Best LLMs for JSON / Tool use in 2026. Ranked by BFCL benchmark, with native JSON output and structured output support.
Verdict
Use Gemini 2.5 Flash for JSON / Tool use today.
GPT-4.1 is the runner-up, 2 points back on BFCL.
2ndShortlist
Researched 15d agoGPT-4.1
- BFCL
- 53.96%
- Output (from)
- $8.00 / 1M
3rdShortlist
Researched 27d agoGranite 4.2 8B
- BFCL
- 52.39%
- Output (from)
- $0.250 / 1M
How we rank
JSON / Tool use leaders rank on BFCL first: every model with a BFCL score ranks above every model without one. Models without BFCL sort by τ-bench. The two benchmarks are never compared with each other.
- Eligibility — Models listed with function calling or tool use.
- Primary ranking — BFCL score first, newer release on ties; models without BFCL follow by τ-bench, then newer release.
- Variant collapse — We keep one row per model family (`familySlug` + parameter tier). When headline scores tie within ±0.5 pt (±10 Elo on Chatbot Arena), we pick the canonical SKU by lowest tracked input price, then GA over preview or limited access, then newest `release`. A folded sibling within the benchmark noise band can show a "Tied within margin" chip on that score cell.
- Podium eligibility — Preview, restricted, retired and retiring models never fill a podium slot. Preview means preview or invite-only access, or a model that is announced or pre-release. Restricted means restricted, partner, private or otherwise unavailable access. Retired means the model is past its retirement date or every provider route is retired. Retiring means the model, or any one of its provider routes, has a retirement date within the next 60 days. These models can still appear in the table below the podium.
- Pricing — Lowest tracked input(output) for API routes.
New models awaiting benchmark coverage
These source-backed rows qualify for this task page, but they are not scored leaderboard picks until the category benchmark data exists.
| Model | Why it is listed | Status | Listed price |
|---|---|---|---|
| Particle-1 Tools | Particle-1 reports tool-use capability; keep it separate from the scored tool-use ranking until benchmarks land. | Benchmark pending No BFCL or tau-bench score listed. | In $0.25 / Out $1.00 |
| LFM3-1.7B Tools | LFM3-1.7B reports tool-use capability; keep it separate from the scored tool-use ranking until benchmarks land. | Benchmark pending No BFCL or tau-bench score listed. | In $0.08 / Out $0.20 |
| GPT-6.1 Terra ToolsCode execution | GPT-6.1 Terra reports tool-use capability; keep it separate from the scored tool-use ranking until benchmarks land. | Benchmark pending No BFCL or tau-bench score listed. | In $1.00 / Out $5.00 |
| Kimi K3.1 Pro Tools | Kimi K3.1 Pro reports tool-use capability; keep it separate from the scored tool-use ranking until benchmarks land. | Benchmark pending No BFCL or tau-bench score listed. | In $2.00 / Out $10.00 |
| # | Model | Input $/1M | Output $/1M | |
|---|---|---|---|---|
| 1 | Claude Opus 4.5 ReasoningVisionTools BFCL (primary): 77.47% | $5.00 | $25.00 | |
| 2 | Claude Sonnet 4.5 ReasoningVisionTools BFCL (primary): 73.24% | $3.00 | $15.00 | |
| 3 | Claude Haiku 4.5 VisionTools BFCL (primary): 68.7% | $1.00 | $5.00 | |
| 4 | Kolibri 1 ReasoningTools BFCL (primary): 61.4% | — | — | |
| 5 | Granite 4.2 30B ReasoningTools BFCL (primary): 61.39% | — | — | |
| 6 | Command A Reasoning (08-2025) Tools BFCL (primary): 57.06% | — | — | |
| 7 | Gemini 2.5 Flash VisionTools BFCL (primary): 56.24% | $0.30 | $2.50 | |
| 8 | GPT-5 Mini ReasoningVisionToolsVerify pricing BFCL (primary): 55.46% | $0.25 | $2.00 | |
| 9 | GPT-4.1 VisionTools BFCL (primary): 53.96% | $2.00 | $8.00 | |
| 10 | Granite 4.2 3B ReasoningTools BFCL (primary): 52.41% | — | — | |
| 11 | Granite 4.2 8B ReasoningTools BFCL (primary): 52.39% | $0.06 | $0.25 | |
| 12 | GPT-5 Nano ReasoningVisionToolsVerify pricing BFCL (primary): 51.45% | $0.05 | $0.40 | |
| 13 | GPT-4.1 Mini VisionTools BFCL (primary): 50.45% | $0.40 | $1.60 | |
| 14 | Kimi K2.5 VisionToolsVerify pricing BFCL (primary): 47.1% | $0.44 | $2.00 | |
| 15 | Command A Tools BFCL (primary): 46.49% | $2.50 | $10.00 | |
| 16 | Mistral Large 2.1 (2411) Tools BFCL (primary): 38.37% | — | — | |
| 17 | Mistral Medium 3 VisionTools BFCL (primary): 37.69% | $0.40 | $2.00 | |
| 18 | Mistral Small 3.2 24B VisionToolsVerify pricing BFCL (primary): 37.15% | $0.10 | $0.30 | |
| 19 | Gemini 2.5 Flash Lite VisionTools BFCL (primary): 36.87% | $0.10 | $0.40 | |
| 20 | Command R7B (12-2024) Tools BFCL (primary): 32.07% | $0.04 | $0.15 |
Honorable mentions
Next seats in this ranking. Lines below are from each model's stored description in LLMReference seed data—spot-check the model page before relying on a capability claim.
Fastest, cheapest GPT-5 variant for summarization and classification tasks. Also available via Realtime API.
51.45%
BFCL
Fast and efficient small model from OpenAI replacing GPT-4o mini. Released April 2025 alongside GPT-4.1. Shows improvements in instruction-following, coding, and intelligence with a 1 million token context window. Available in ChatGPT for paid users.
50.45%
BFCL
Kimi K2.5 is Moonshot AI's Kimi model focused on code generation and software engineering. It offers a 256K-token context window and scores 87.9 on GPQA.
47.1%
BFCL
Compare Top Picks
Side-by-side comparison of the top picks by price, benchmark, and API access.
Browse Other Categories
Best LLMs for Code GenerationBest LLMs for RAGBest LLMs for Agents vs Agent Products (2026)Best LLMs for ClassificationBest Open Source LLMsBest Multimodal / Vision LLMsBest LLM for Translation in 2026Best AI Image Models in 2026Best AI Video Models in 2026Best LLMs for Reasoning & MathBest Small Language Models (SLMs)Cheapest LLM APIs You Can Call Right NowBest Long Context LLMsBest Mainstream LLM APIs, RankedBest LLMs for EnterpriseBest Free LLMs You Can Use Right NowBest LLMs for WritingBest LLMs for MarketingBest LLMs for Customer Support
Frequently asked questions
What is the best LLM for function calling and tool use?
LLM Reference's top pick for function calling and tool use as of 2026-10-29 is Gemini 2.5 Flash by Google DeepMind, with BFCL 56.24%. Picks come from models that meet the podium eligibility rules, so the top pick may not lead the full table. Second pick: GPT-4.1 by OpenAI (BFCL 53.96%).
Which ranked LLM for function calling and tool use has the lowest output price?
Among the ranked models shown on this page, Command R7B (12-2024) by Cohere has the lowest listed output price: $0.150/1M output tokens, at rank 20.
What is the best open-weight LLM for function calling and tool use?
Kolibri 1 by Aleph Alpha is the highest-ranked open-weight model shown on this page (rank 4, BFCL 61.4%). License: Apache 2.0.
How is the function calling and tool use ranking calculated?
JSON / Tool use leaders rank on BFCL first: every model with a BFCL score ranks above every model without one. Models without BFCL sort by τ-bench. The two benchmarks are never compared with each other. Preview, restricted, retired and retiring models never fill a podium slot. Preview means preview or invite-only access, or a model that is announced or pre-release. Restricted means restricted, partner, private or otherwise unavailable access. Retired means the model is past its retirement date or every provider route is retired. Retiring means the model, or any one of its provider routes, has a retirement date within the next 60 days. These models can still appear in the table below the podium.
How often is this page updated?
Weekly. Last refreshed 2026-10-29.