xpu liveBETA
Models›Motif 3

Motif 3

Model reference ↗

Model ranking

Not matched to the current verified ranking snapshot.

Pricing 0 Model information

Provider prices 0

Lowest input price per billing unit. Conditions in Details.

ProviderPriceProvider model IDStatusContext / max outputSource

Motif 3 is developed by Motif Technologies. This profile brings together its verified specifications, access information and independent benchmark reference so you can evaluate the model before choosing a deployment.

Architecture and context

Motif 3 is a sparse mixture-of-experts language model from Motif Technologies. Its developer reports 314 billion total parameters, with 13.2 billion active for each token, and a native 262,144-token context window. The architecture uses 384 routed experts, eight selected per token, and one shared expert.

Training and intended use

The model card describes approximately 12.5 trillion training tokens spanning English, Korean, code and mathematics, with additional emphasis on legal and financial material. This makes bilingual document work and technical reasoning relevant evaluation tasks. Domain emphasis does not replace checking the accuracy of a legal or financial answer.

Weights and deployment

Motif 3 weights are available under the MIT license. The main checkpoint uses BF16; a separately published NVFP4 variant has different deployment requirements. The developer provides a custom vLLM path and Motif-specific reasoning and tool parsers. Check the selected checkpoint and serving recipe before planning infrastructure.

Comparing Motif 3 on XPU

Use the model rankings for the catalog’s current intelligence score and its checked date. Scores describe the tested configuration; a similar name or a different reasoning setting may produce different results. Compare this profile with the model rankings and review provider information before selecting an API.

Rental and token prices depend on the provider, region, routing and deployment. An empty offers table means XPU has no matching verified offer at present. It is not a zero-price quote. Check provider limits separately from the developer’s maximum context specification.

Sources and verification

Motif 3 — official developer model card. Specifications reviewed on 18 October 2026. The image represents the developer’s brand, rather than a separate model-specific logo. Unknown details have been left unconfirmed instead of estimated.

CONTEXT TOKENS262,144Developer model card
MAX OUTPUT TOKENSNot suppliedProvider-dependent
PROVIDERS0Linked in this catalog
SOURCE CHECKED2026-10-18Sources reviewed; unknown fields labeled

Specifications

Catalog model ID
motif/motif-3
Developer model ID / revision
Motif-3
Input modalities
textDeveloper model card ↗
Output modalities
textDeveloper model card ↗
Context / input token limit
262,144Developer model card ↗
Maximum output tokens
Not verified in cited sources
Tool calling
SupportedDeveloper model card ↗
Structured output
Not verified in cited sources
API parameters (source-specific)
Not verified in cited sources
Weight access
Open weightsDeveloper model card ↗
Weight license / access terms
MITDeveloper model card ↗
Total parameters (billions)
314Developer model card ↗
Active parameters (billions)
13.2Developer model card ↗
Weight format / quantization
BF16; separate NVFP4 checkpointDeveloper model card ↗
Download size (GB)
Depends on checkpoint and precision; no verified file size
Release date
Not verified in cited sources
Notes
Developer specifications may differ from hosted endpoint limits. Parameter counts follow the developer model card; weight-file metadata may report a different counting convention.

Catalog entries link to their specification and availability sources. This is a curated selection, not every model each provider offers. Best prices are the lowest collected USD tariffs, not a total-cost estimate. Token offers are selected by input price, with output from the same tariff; workload mix may change the cheapest provider. Listed variants may require batch processing, off-peak hours, or specific media settings. Price coverage varies by provider; collected tariffs include their source and observation time. Arena user-preference scores are checked automatically every six hours against the latest published Text / overall dataset. Source: Arena / LMArena; CC BY 4.0. Selection and catalog matching by XPU LIVE. They apply to the shown evaluation configuration; listed prices may use different settings. Models without verified scores are not assumed to be weaker.