- GLM Legal pricing has no public rate card; commercial access runs through official consultation
- Open weights for
law-glm-10bandlaw-glm-10b-v2are free under Apache-2.0 on Hugging Face - Self-hosting costs come from GPU compute, not licensing fees
- Enterprise access requires the official legal consultation channel to get a quote
- No hidden tiers are confirmed; beware of third-party sites claiming fixed prices
GLM Legal Pricing: Current Status in 2026
GLM Legal pricing is not published as a fixed public rate card. As of October 2026, access to the GLM-Legal product line is handled through an official consultation process, while the related open-weight models law-glm-10b and law-glm-10b-v2 are available at no charge on Hugging Face under the Apache-2.0 license.
That means the honest answer to "how much does GLM Legal cost" depends on which access path you take:
| Access Path | Public Price | Availability | Best For |
|---|---|---|---|
| Open weights (law-glm-10b / v2) | Free download | Public on Hugging Face | Researchers, self-hosting teams |
| Official GLM-Legal consultation | Quote-based | Via official consultation entry | Law firms, enterprises |
| GLM Cookbook development resources | Free | Public on GitHub | Prototype builders |
| Third-party hosted APIs | Set by provider | Varies, unofficial | Experimentation only |
No official GLM Legal pricing tiers, monthly subscriptions, or per-token rates have been publicly confirmed. Treat any website publishing exact dollar figures for "GLM Legal plans" with caution unless the numbers come directly from an official channel.
Free Access: The Open-Weight Models
The cheapest way to use GLM Legal technology is downloading the open model weights yourself. The law-llm organization on Hugging Face hosts two model variants:
| Model | Parameters | License | File Size | Notes |
|---|---|---|---|---|
| law-glm-10b | 10B | Apache-2.0 | ~39.5 GB | Base legal-domain GLM model |
| law-glm-10b-v2 | 10B | Apache-2.0 | Similar | Updated variant of the base model |
Because both are released under Apache-2.0, you can use, modify, and commercially deploy the weights without paying a license fee, subject to the license's conditions. The THUDM GLM repository provides the upstream code and architecture background.
The license costs nothing, but running a 10B-parameter model still requires GPU hardware or rented cloud compute. Budget for VRAM, storage, and electricity when comparing self-hosting against consultation-based access.
Consultation-Based Enterprise Pricing
For organizations that want the GLM-Legal product rather than self-hosted research weights, pricing is determined individually through the official Z.ai's official site. A quote typically reflects your deployment model, expected volume, and integration requirements.
Define your legal workflow
Document which tasks you need support for — contract review, evidence analysis, legal research, compliance, or drafting — and the jurisdictions involved.
Estimate volume and load
Prepare expected document counts, request traffic, latency needs, and concurrency requirements. These directly affect quoted pricing.
Specify data-handling requirements
List privacy, retention, and access-control obligations. Regulated data often changes deployment options and cost.
Request a consultation
Contact the official consultation entry with the prepared information and ask for a scoped proposal covering API access, service scope, and commercial terms.
GLM-Legal as an official product is distinct from the LawGLM open-source research project and the law-glm-10b weights. Confirm exactly which offering a quote covers before signing anything.
Self-Hosting Cost Factors
If you choose the free Apache-2.0 weights, your real costs shift to infrastructure and operations. The main budget categories:
Compute
- 10B model needs substantial VRAM
- GPU rental or owned hardware
- Inference speed vs. cost tradeoff
Engineering
- Setup with Transformers/PyTorch
trust_remote_codeconfiguration- Ongoing maintenance and updates
Compliance
- Data privacy safeguards
- Human review workflow
- Audit logging and monitoring
| Cost Category | Fixed or Variable | Typical Driver |
|---|---|---|
| GPU hardware or rental | Variable | Usage hours, batch size |
| Model storage | Fixed | ~39.5 GB per variant |
| Engineering time | Fixed | Setup, prompt tuning, evaluation |
| Legal review overhead | Variable | Output volume requiring sign-off |
With Apache-2.0, there are no per-seat or per-token fees owed to the model publisher. Your self-hosting spend is entirely infrastructure and labor, which makes cost prediction more straightforward than opaque subscription tiers.
Pre-Consultation Checklist
Prepare these items before requesting a GLM Legal enterprise quote so the consultation can produce a concrete price:
Before You Ask for a Quote:
- Document primary legal workflows and objectives
- List applicable jurisdictions and languages
- Estimate document volume, formats, and corpus size
- Define privacy, retention, and access-control needs
- Specify human review and audit requirements
- Decide between cloud API or private deployment
GLM Legal Pricing FAQ
Q: Is GLM Legal free to use?
The open model weights law-glm-10b and law-glm-10b-v2 are free to download and use under the Apache-2.0 license on Hugging Face. The GLM-Legal commercial product is priced individually through official consultation rather than a public rate card.
Q: Where can I find an official GLM Legal price list?
No public price list or subscription tier has been confirmed as of October 2026. Organizations should use the official legal consultation channel to receive a scoped commercial proposal.
Q: Can I use the free weights commercially?
The Hugging Face model pages list Apache-2.0, a permissive license that generally allows commercial use, modification, and redistribution subject to its conditions. Review the license text and verify any additional terms on the model page before deploying.
Q: What drives the cost of self-hosting GLM Legal?
The main costs are GPU compute for a 10B-parameter model, roughly 39.5 GB of storage per variant, engineering time for setup and prompt tuning, and compliance overhead such as human review and audit logging.
Always confirm pricing and terms through official channels. This wiki tracks publicly available information and does not set or guarantee any prices for GLM Legal.