Writer

“WRITER is where the world's leading enterprises orchestrate AI-powered work.” — the vendor’s own words

A real full-stack enterprise agent platform with a genuine MCP-based connector spine and a 1M-token model — the catch is the price tag: no mid-tier, and Enterprise pricing plus deployment effort put it out of reach for anyone below a few hundred seats.

The verdict, read by Jan · AI clone of Jan’s voice, generated locally with his consent.

Best for: Large enterprises (regulated industries especially — healthcare, finance) that want one governed platform for agentic workflows across teams and are prepared to go through a sales-led Enterprise deployment rather than self-serve.

Scope17/20

Quality8/10

Where the quality sits

8Reactive

6Retrieval & Memory

9Orchestration

7Validation

8Models

Enterprise platformAutomation & AgentsCopywritingProductivityPaid

Vendor

Writer, Inc. · writer.com

Origin

US — San Francisco, CA

Pricing

Starter ~$29–39/user/month (third-party reported) · Enterprise Custom

Users (official only)

Not disclosed precisely. Writer's own newsroom/about page says 'hundreds of the world's leading enterprises' and 'hundreds of customers'; third-party press (referenced in Writer's own newsroom) cites '300 companies' as of April 2025. (source, 2026-08-03)

PRICING

Starter	~$29–39/user/month (third-party reported)	Up to 5 users, 5 Playbooks, basic connectors, limited Knowledge Graph, Basic guardrails, 14-day free trial
Enterprise	Custom	Unlimited users/playbooks, full Knowledge Graph, advanced orchestration/approvals, granular governance/observability/auditability, SAML SSO, SCIM, SOC 2 Type II report, HIPAA BAA available

Writer's live pricing page (writer.com/pricing) currently lists feature comparisons but no numeric Starter price; the $29-39/user/month figure comes from third-party trackers (G2, TrustRadius, eesel.ai) checked 03/08/2026.

checked 2026-08-03 · vendor pricing page

Element scores

Reactive

Retrieval & Memory

Orchestration

Validation

Models

Primitives

Compositions

Deployment

Emerging

Tap or hover any element to see why it got that score.

Strengths

A genuinely full-stack enterprise agent platform: Palmyra X5's 1M-token context window (now Writer's stated minimum for all future models), a graph-based RAG layer (Knowledge Graph) that outperformed 7 vector-based RAG approaches on Writer's own RobustQA benchmark, and a connector spine built natively on MCP with a dedicated engineering security write-up — plus real domain fine-tuning (Palmyra-Med-70B, Palmyra-Fin-70B) for regulated industries.

Honest dings

No mid-tier plan — it's 5-user Starter or full custom Enterprise, with governance, observability and auditability locked behind Enterprise. No standalone embeddings API or vector store (Knowledge Graph is graph-only by design), no documented customer-facing eval/regression-testing suite, and no customer-exposed interpretability surface. Official pricing page has stopped publishing numbers at all, making self-serve cost comparison harder than it used to be.

Prices and details change — this passport is re-verified at least quarterly.

Sources (16) — every claim traceable

Every audit lists the research it rests on — transparency and traceability are the product. Tools evolve: each audit is a snapshot of its audit date, and re-audits supersede older versions (kept below for reference).

writer.com/pricing — Official plans/features comparison table (Starter vs Enterprise); no numeric Starter price currently listed (accessed 2026-08-03)
writer.com/product/connectors — Official confirmation that Writer's connector infrastructure is built natively on MCP, including a purpose-built MCP gateway (accessed 2026-08-03)
writer.com/engineering/mcp-security-considerations — Writer engineering blog (Director of Engineering) on MCP security risks and mitigations (accessed 2026-08-03)
writer.com/product/graph-based-rag — Official Knowledge Graph RAG product page, RobustQA benchmark vs 7 vector RAG approaches (accessed 2026-08-03)
writer.com/llms/palmyra-x5 — Official Palmyra X5 model page: 1M context, BBH/GPQA/MMLU-Pro benchmark scores, ~22s/1M-token processing, ~300ms tool-call latency (accessed 2026-08-03)
aws.amazon.com/about-aws/whats-new/2025/04/writers-… — AWS Bedrock listing confirming Palmyra X5 (1M context) and X4 (128K context) availability (accessed 2026-08-03)
sacra.com/c/writer — Independent revenue/funding analysis: $47M ARR (Nov 2024), $326M total funding, $1.9B valuation, Palmyra parameter range 128M-43B (accessed 2026-08-03)
writer.com/blog/series-c-funding-writer-press-relea… — Official Series C press release: $200M at $1.9B valuation, customer list (Mars, Ally Bank, Salesforce, Uber) (accessed 2026-08-03)
writer.com/about — Official About page: founders, HQ address, 'hundreds of leading enterprises' customer language (accessed 2026-08-03)
writer.com/newsroom — Official newsroom listing press coverage including '300 Companies' reference and hundreds-of-customers language (accessed 2026-08-03)
baseten.co/resources/customers/writer — Case study confirming Palmyra-Med-70B and Palmyra-Fin-70B as custom domain-specific fine-tuned models (accessed 2026-08-03)
dev.writer.com/llms.txt — Official developer docs index: Agent Builder observability, session logs, MCP-based Databricks connector, Applications API (accessed 2026-08-03)
dev.writer.com/home/custom-connectors — Official docs: custom connectors buildable from OpenAPI specs or MCP servers (accessed 2026-08-03)
writer.com/blog/writer-agent-skills-playbooks-press… — Official press release: 5,000+ playbooks deployed by customers, 333% ROI per Forrester TEI study (accessed 2026-08-03)
eesel.ai/blog/writer-com-pricing — Independent pricing analysis citing $29-39/user/month Starter tier and 300+ enterprise customers (accessed 2026-08-03)
vendr.com/marketplace/writer — Independent contract-value data: median $28,910/year, range $12,916-$101,285 (accessed 2026-08-03)