Legora

Legal task quality: 6/10

Estimate from published benchmarks · Assessed 25 October 2026

Estimated from the assumed model's results on legal research and drafting tasks under US law. The benchmark does not test Finnish or EU law, or the tool's own sources and workflows.¹⁴

Assumption: Legora names its AI model providers but not the models, and no published benchmark covers Legora. We therefore assume a low-cost general-purpose model and score OpenAI's GPT-6 Luna, the low-cost model of OpenAI's current generation, which all three benchmarks we use list. A stronger model would score higher; we replace the estimate when the model is disclosed or the tool is benchmarked.

Grounding and verifiability: 3/10

Estimate from published benchmarks · Assessed 25 October 2026

Estimated from how often the assumed model's research answers meet every point of an expert grading rubric, sources included. The tool's own source library and safeguards may do better than the model alone, which no published benchmark measures yet.¹⁴

Assumption: Legora names its AI model providers but not the models, and no published benchmark covers Legora. We therefore assume a low-cost general-purpose model and score OpenAI's GPT-6 Luna, the low-cost model of OpenAI's current generation, which all three benchmarks we use list. A stronger model would score higher; we replace the estimate when the model is disclosed or the tool is benchmarked.

Finnish quality: 6/10

Estimate from published benchmarks · Assessed 27 October 2026

Estimated from the assumed model's general Finnish results. The benchmark does not test legal terminology.¹⁴

Assumption: Legora names its AI model providers but not the models, and no published benchmark covers Legora. We therefore assume a low-cost general-purpose model and score OpenAI's GPT-6 Luna, the low-cost model of OpenAI's current generation, which all three benchmarks we use list. A stronger model would score higher; we replace the estimate when the model is disclosed or the tool is benchmarked.

Finnish legal sources: 3/10

From public documentation · Assessed 27 October 2026

Legora lists Finnish legislation, case law and government proposals through Alma's Suomen Laki Service, but documents no links between them that its AI uses.^22,42

EU legal sources: Not yet scored (5 to 10)

From public documentation · Assessed 27 October 2026

Not yet scored: we could not check whether Legora covers EU case law, because its help-centre article on legal sources needs a login. Its public pages list EUR-Lex.^5,22,43

Security and data handling: 9/10

From public documentation · Assessed 27 October 2026

Legora hosts each workspace in one region, the EU among them, and every AI model provider in the EU section of its subprocessor list processes in the EU/EEA. Its EU terms rule out training on customer content and say it deletes content 30 days after the agreement ends, its EU DPA is public, and it lists SOC 2 Type II, ISO/IEC 27001 and SAML single sign-on.^{14,21,40,41,44,46–51}

Workflow fit: 10/10

From public documentation · Assessed 23 October 2026

Legora has Word and Outlook add-ins, connects to iManage and NetDocuments, runs an agent and no-code workflows, provisions users through Directory Sync, keeps audit trails of agent actions, and connects to MCP servers, including ones built on an organisation's own systems.^{10,15,19,20,25–27,52}

Estimated price: 3/10

From an estimated price · Assessed 27 October 2026

Legora publishes no prices. A buyer's guide by Layer3Labs reports about US$3,000 per user per year, about US$250 a month, with about 10 seats or more.^63,64

Price we score: US$250 per user per month, excluding VAT, about €223 at the European Central Bank reference rate of 23 October 2026.

Assumption: We use the buyer's guide figure, which does not name its source. Sacra reports an average contract of about US$95,000 a year per organisation, which gives no price per user.