Research / developmentPartially privateHigh confidence
Argonne Leadership Computing Facility Inference Endpoints
No-cost research and development inference across 7 models; not unrestricted production use.
Free accessResearch / development
Payment cardUnknown
AccountNot documented
Sources7 first-party links
API endpoint
https://inference-api.alcf.anl.govModels mentioned
meta-llama/Meta-Llama-3.1-405B-Instructmeta-llama/Llama-4-Maverick-17B-128E-Instructmistralai/Mistral-Large-Instruct-2407openai/gpt-oss-120bargonne/AuroraGPT-IT-v4-0125google/gemma-4-31B-itnvidia/nemotron-3-super-120bEquivalent paid value
How this was valued: The published allowance or a defensible paid comparison is not precise enough to calculate.
User allocation is project-specific and the facility has no like-for-like commercial unit price.
Limits and terms
- Numeric User Limit
- allocation specific
- Batch Max Requests
- 150,000
- Cold Start Minutes
- 10-15
What happens to your prompts?
Partially privatePartial review
Institutional controls apply, but endpoint-wide content retention, training, administrator access, and deletion terms are not public.
- Plan Scope
- Argonne Leadership Computing Facility research allocations and inference endpoints; project and Department of Energy policies apply.
- Prompt Retention
- No endpoint-wide public prompt-retention TTL was found.
- Response Retention
- not documented
- Ordinary Logging
- Facility authentication, allocation, security, and operational telemetry may be logged; content logging is not documented in the endpoint guide.
- Model Training
- No public commitment excludes research workload content from secondary analysis or model training across all projects.
- Product Improvement
- Facility telemetry may support operations and research; content scope is not specified.
- Human Or Operator Access
- Facility administrators and project personnel may have privileged technical access under institutional controls.
- Subprocessors And Routing
- Workloads run on Argonne/ALCF systems and are subject to allocation and project governance rather than consumer SaaS terms.
- Deletion Controls
- Project storage and account controls apply; no inference-request deletion workflow is documented.
- Caveat
- Institutional access approval is not a privacy guarantee. Users should rely on their allocation agreement and data-management plan before sending controlled data.
No provider terms or privacy-policy link is captured in this snapshot. Review the provider’s current legal documents before sending sensitive data.
Eligibility
- Approved Project Required
- Yes
- Programs
- Directors DiscretionaryINCITEALCCNAIRR
- Open Research Cost
- generally no cost compute allocation
- Proprietary Research
- cost recovery
Primary sources
Official DocsInference endpointshttps://docs.alcf.anl.gov/services/inference-endpoints/Official DocsAllocation managementhttps://docs.alcf.anl.gov/account-project-management/allocation-management/Official ProgramDiscretionary allocationshttps://www.alcf.anl.gov/science/directors-discretionary-allocation-programOfficial RepositoryInference endpoints repositoryhttps://github.com/argonne-lcf/inference-endpointsOfficial Annual Report2025 ALCF Annual Report, About ALCF (2,103 users, 483 active projects, 274 publications)https://ar25.alcf.anl.gov/year-in-review/about-alcfOfficial Repositoryargonne-lcf GitHub organizationhttps://github.com/argonne-lcfOfficial ProductAurora exascale supercomputer at ALCFhttps://www.alcf.anl.gov/aurora