LLM Reference

MT0 Models by BigScience

BigScienceApache 2.0Open source
5 models2024Up to 1k ctxFrom $1.8/1M input

Last refreshed 2026-05-19. Next refresh: weekly.

Details

ResearcherBigScience
LicenseApache 2.0OSI-approved
Commercial useCommercial use: permitted
Models5
Released2024
Max context1k

About

MT0 is a family of 5 AI models by BigScience, released in 2024.

Decision facts

Best fit
codingstructured outputs
Capability starting point
MT0 XXL with 1k context
Lowest tracked input
MT0 XXL · $1.8/1M · IBM watsonx
Closest related family
Claude 3

Current Variants

Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.

5 in view
MT0 XXLCurrent

Use when the workload needs 1k context and 13B parameters.

2024-011k context13B parameters
MT0 XLCurrent

Use when the workload needs 1k context and 3.7B parameters.

2024-011k context3.7B parameters
MT0 LargeCurrent

Use when the workload needs 1k context and 1.2B parameters.

2024-011k context1.2B parameters
MT0 BaseCurrent

Use when the workload needs 1k context and 580M parameters.

2024-011k context580M parameters
MT0 SmallCurrent

Use when the workload needs 1k context and 300M parameters.

2024-011k context300M parameters

Release Timeline

1 release group
2024-01
5 current
MT0 Base
1k context580M parameters
Current
MT0 Large
1k context1.2B parameters
Current
MT0 Small
1k context300M parameters
Current
MT0 XL
1k context3.7B parameters
Current
MT0 XXL
1k context13B parameters
Current

Specifications(5 models)

MT0 model specifications comparison
ModelReleasedContextParameters
MT0 XXL2024-011k13B
MT0 XL2024-011k3.7B
MT0 Large2024-011k1.2B
MT0 Base2024-011k580M
MT0 Small2024-011k300M

Available From(1 provider)

Pricing

MT0 model pricing by provider
ModelProviderInput / 1MOutput / 1MType
MT0 XXLIBM watsonx$1.8$1.8Serverless