Skip to content

AI21 Labs · Chat

Jamba 1.5 Mini

AI21 Labs 🇮🇱ChatJamba familyThe maker currently plans to deprecate it within the next 180 days (26 Nov 2026). Plans can change.
llminfo IQ
98.6
Maker's price
$0.25 blended / 1M
Max context
256K
Listings
8

Evidence and sources

llminfo IQ 98.6, ranked among chat models. From 113 benchmarks. Plenty of data behind it.How llminfo IQ is computedPoor value (26.9). Costs more per task than its benchmark results predict. From 3 benchmarks with measured task costs. A fair amount of data behind it.How Value for money worksMaker HQ: IL (according to AI21 Labs's own website)
Sources: LMArena data, LiteLLM dataPrimary data from LMArena from LiteLLM: deprecation date

Advantages and trade-offs

Compared with 10 chat models, picked mainly by popularity and IQ

Trade-offs

Deprecation scheduled

The maker plans to deprecate it on 26 Nov 2026, within the next 180 days.

We only make a claim when there's enough data on models of the same type. Updated 22 Nov, 04:34 UTC.Prices come from the model's maker where they sell it, otherwise their official partner, otherwise the market median. This comparison uses 7 maker's prices; 3 market median prices.

Input & output

as stated by the model's providers
Text
Text

Listings

 AI21 Labsjamba-1.5-miniSource: LiteLLM data256K$0.20$0.40
 AI21 Labsjamba-1.5-mini@001 · 001Source: LiteLLM data256K$0.20$0.40
 AI21 Labsjamba-mini-1.5-2024-08 · 2024-08Source: Portkey Models dataNot published by the provider$0.20$0.40
 Googlevertex_ai/jamba-1.5-miniSource: LiteLLM data256K$0.20$0.40
 Googlevertex_ai/jamba-1.5-mini@001 · 001Source: LiteLLM data256K$0.20$0.40
 Flowdotjamba-1.5-miniSource: FlowDot data256K$0.20$0.40
 OpenRouterai21/jamba-1-5-miniSource: GenAI Prices dataNot published by the provider$0.20$0.40

1 listing without a price is hidden.

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Knowledge

HELM59 related benchmarks · 59 results
Show 53 more resultsHide 53 results
MMLUSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
MMLU-ProSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Human preference

Arena24 related benchmarks · 24 results
ELO
Source: LMArena data · Nov 2026
ELO - Chinese
Source: LMArena data · Nov 2026
ELO - Creative Writing
Source: LMArena data · Nov 2026
ELO - English
Source: LMArena data · Nov 2026
ELO - Exclude Ties
Source: LMArena data · Nov 2026
ELO - Expert
Source: LMArena data · Nov 2026
Show 18 more resultsHide 18 results
ELO - Japanese
Source: LMArena data · Nov 2026
ELO - Longer Query
Source: LMArena data · Nov 2026
ELO - Multi Turn
Source: LMArena data · Nov 2026
ELO - Non English
Source: LMArena data · Nov 2026
ELO - Russian
Source: LMArena data · Nov 2026
ELO Style-Controlled
Source: LMArena data · Nov 2026
Hard
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Long context

Reasoning

ARC-CSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Math

GSM8KSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Science

GPQASource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim

Factuality

General

Coding

Instruction following

Multilingual

Other

ValsSource: Vals AI

Data credits: Data by LLM Stats · LMArena, CC BY 4.0 (licence) · Stanford HELM (Stanford CRFM), Apache-2.0 (licence) · Stanford HELM (Stanford CRFM); scenario scores from Capabilities, Safety, Long Context, MedHELM, AIR-Bench, MMLU and Finance · NVIDIA RULER, Hsieh et al. Official README numeric table; no separate results licence asserted.

Around the web

  • Jamba: A Hybrid Transformer-Mamba Language Model | AI21
    Transformer and Mamba layers, and how to mix experts, and show that some of them …Transformer and Mamba layers, and how to mix experts, and show that some of them are crucial in large scale modeling. We also describe several interesting properties of these architectures which the …

    ai21.com

  • Jamba-1.5: Hybrid Transformer-Mamba Models at Scale | AI21
    results while providing high throughput and outperforming other open-weight models on long-context benchmarks. The model …results while providing high throughput and outperforming other open-weight models on long-context benchmarks. The model weights for both sizes are publicly available under the Jamba Open Model Licen…

    ai21.com

  • configuration_jamba.py · ai21labs/Jamba-v0.1 at main
    Follow AI21 642 Text Generation Transformers Safetensors jamba mamba Mixture of Experts custom_code arxiv: 2403.19887 …Follow AI21 642 Text Generation Transformers Safetensors jamba mamba Mixture of Experts custom_code arxiv: 2403.19887 License: apache-2.0 Model card Files Files and versions xet Community 51 Deploy C…

    huggingface.co

  • README.md · ai21labs/AI21-Jamba-Mini-1.5 at main
    Jamba-Mini-1.5 like 271 Follow AI21 639 Text Generation Transformers Safetensors 9 languages jamba mamba Mixture …Jamba-Mini-1.5 like 271 Follow AI21 639 Text Generation Transformers Safetensors 9 languages jamba mamba Mixture of Experts conversational arxiv: 2305.14314 License: jamba-open-model-license Model ca…

    huggingface.co

  • README.md · ai21labs/AI21-Jamba-Large-1.5 at main
    Run the training process by using the [example](https://github.com/axolotl-ai-cloud/axolotl/blob/main/examples/jamba/qlora_fsdp_large.yaml) (change the `base_model` to the pre-quantized model …Run the training process by using the [example](https://github.com/axolotl-ai-cloud/axolotl/blob/main/examples/jamba/qlora_fsdp_large.yaml) (change the `base_model` to the pre-quantized model or to `…

    huggingface.co

  • ai21labs (AI21)
    45 Jamba Reasoning 3B AI21's top-performing reasoning model that packs leading scores on intelligence benchmarks …45 Jamba Reasoning 3B AI21's top-performing reasoning model that packs leading scores on intelligence benchmarks and highly-efficient processing into a compact 3B build ai21labs/AI21-Jamba-Reasoning-…

    huggingface.co

  • ai21labs/Jamba-v0.1 · Fast Mamba

    AutoTokenizer.from_pretrained("ai21labs/Jamba-v0.1", trust_remote_code=True) model = AutoModelForCausalLM.from_pretrained("ai21labs/Jamba-v0.1", trust_remote_code=True, device_map="auto") Notebooks G…

    huggingface.co

  • ai21labs/AI21-Jamba2-Mini · Hugging Face
    ai21labs / AI21-Jamba2-Mini like 56 Follow AI21 643 Text Generation Transformers Safetensors jamba conversational License: …ai21labs / AI21-Jamba2-Mini like 56 Follow AI21 643 Text Generation Transformers Safetensors jamba conversational License: apache-2.0 Model card Files Files and versions xet Community 2 Deploy Copy t…

    huggingface.co

  • ai21labs/AI21-Jamba2-3B · Hugging Face
    Jamba2 3B is an ultra-compact open source model designed to bring enterprise-grade reliability to on-device …Jamba2 3B is an ultra-compact open source model designed to bring enterprise-grade reliability to on-device deployments. At just 3B parameters, it runs efficiently on consumer devices—iPhones, Androi…

    huggingface.co

  • ai21labs/Jamba-v0.1 · Just a solid congrats and thank you to your team

    tokenizer = AutoTokenizer.from_pretrained("ai21labs/Jamba-v0.1", trust_remote_code=True) model = AutoModelForCausalLM.from_pretrained("ai21labs/Jamba-v0.1", trust_remote_code=True) Notebooks Google C…

    huggingface.co

Also known as: ai21.jamba-1-5-mini-v1:0, ai21/jamba-1-5-mini, jamba-1.5-mini@001, jamba-mini-1.5-2024-08, snowflake/jamba-1.5-mini, vertex_ai/jamba-1.5-mini, vertex_ai/jamba-1.5-mini@001

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works