IBM Research
61 models across 10 families · Latest: Granite 4.1 3B (2026-04)
Creating reliable and adaptable AI solutions
IBM Research's portfolio covers 59 active models across 10 current families, spanning coding, rag, and agents. Open a model detail page to compare provider routes and sourced benchmarks.
Covers 7 workload areas across 59 active tracked models; last verified 2026-06-04.
Use it for
- Teams evaluating coding, rag, and agents across this lab's releases
- Comparing model families before committing to a flagship
- Migration and pricing follow-ups across 59 tracked models
Do not use it for
- Choosing a hosting provider without opening a model page for price ladders
Active models
59
Current models from this lab, excluding deprecated ones
Active families
10
Current model families from this lab
Open catalog
59 open
47 open source / 12 open weights
Lowest output price
$0.100 /1M
Cheapest tracked output across active models, per 1M tokens
Latest dated release
2026-04-29
Granite 4.1 3B
Freshness
2026-06-04
Researched 90d ago
Information
Release cadence
- Dated releases
- 5
- Latest model
- Granite 4.1 3B
- Latest date
- 2026-04-29
Where this lab wins
- Coding: 3 tracked models with SWE-bench / HumanEval-style scores.
- RAG: 1 tracked model with ruler / needle retrieval benchmarks.
- Agentic: 4 tracked models with BFCL, tau-bench, and SWE-bench tool-use coverage.
- Long-context: 34 tracked models with context-token or InfiniteBench-class signal.
Flagship quality / price signal
- Flagship
- Granite 4.1 8B
- Selection source
- Best sourced coding Q/$
- Coding grade
- A
- Benchmark / output price
- humaneval 87.2 · $0.100/1M
About
IBM Research, established in 1945 and based in Armonk, New York, has been an influential player in the field of artificial intelligence (AI) for decades, setting benchmarks through various historical milestones. Noteworthy achievements include the development of Arthur Samuel's self-learning checkers program in the 1950s and William Dersh's voice-operated "Shoebox" in 1962. These early innovations paved the way for the development of more sophisticated AI systems, including Deep Blue's famous 1997 chess victory over Garry Kasparov and Watson's win on Jeopardy! in 2011. These milestones are foundational to IBM's ongoing innovations in generative AI and large language models (LLMs).
Featured models
| Model | Released | Context | Input price ($/1M) | Output price ($/1M) | License | Openness |
|---|---|---|---|---|---|---|
| Granite 4.1 3B | 2026-04-29 | 131k | - | - | Apache 2.0 | Open source |
| Granite 4.1 3B Base | 2026-04-29 | 512k | - | - | Apache 2.0 | Open source |
| Granite 4.1 8B | 2026-04-29 | 131k | $0.05 | $0.1 | Apache 2.0 | Open source |
Model families
Recent releases
- Granite 4.1 3B- 2026-04-29
- Granite 4.1 3B Base- 2026-04-29
- Granite 4.1 8B- 2026-04-29
- Granite 4.1 8B Base- 2026-04-29
- Granite 4.1 30B- 2026-04-29
Top comparisons
- Granite Vision 4.1 4B vs Qwen3.5-397B-A17B162
- Together AI - Gemma 3n-e4B vs Granite Guardian 3.0 8B62
- Granite Guardian 3.0 8B vs Llama Guard 7B33
- Granite Guardian 3.0 8B vs Italia 10B Instruct29
- Granite Guardian 3.0 8B vs Llama Guard 2 8B26
- Granite Guardian 3.0 8B vs Llama 3.1 NemoGuard 8B Content Safety21
- GPT-2 Medium vs Granite Guardian 3.0 8B20
- Granite 3.3 8B Instruct vs Trinity-Large-Preview19
Explore related pages
Last reviewed: 2026-06-04. Data sourced from public lab announcements and provider documentation.