Choose the right Palmyra model for your needs. Compare capabilities, context windows, and pricing for the Palmyra X6, X5, and X4 models.
Deprecation notice: Palmyra X4 (palmyra-x4) and Palmyra X5 (palmyra-x5) will be deprecated on December 14, 2026.Migration path: Use Palmyra X6 (palmyra-x6) instead. Palmyra X6 has a 1M token context window (Palmyra X4 is 128k). See pricing for cost details and the deprecation policy for more information.
Whether you use no-code, Agent Builder, or APIs, you need to choose a model. Here is an overview of the Palmyra models and their capabilities.
Palmyra X6 is Writer’s default agentic model for WRITER Agent and long-horizon enterprise workflows. It was co-developed with the product’s orchestration layer. Capability scores in the Benchmarking tab measure the model in that combined system, which is how customers deploy it.Palmyra X6 can hold a single objective for up to 8 hours without supervision, planning, executing, verifying output, and delivering finished work rather than a draft that needs another pass.Pricing:
Input: $2.00 per 1M tokens
Output: $8.00 per 1M tokens
Context window: 1MTask economics (median on Writer’s internal evaluation of production customer tasks, n=22):
Cost per finished task: $0.12
Latency per finished task: 26 seconds
Palmyra X6 is built for the agentic work marketing, revenue, and operations teams run in production, including grounded research, content generation, connector-driven actions, and multi-step playbooks.WRITER Agent workflows
Connector-driven tasks: Query CRM records, search Slack messages, update documents, and perform other actions through prebuilt and custom connectors
Multi-step automation: Break complex requests into sequenced tool calls across multiple systems
Grounded responses: Pull context from Knowledge Graphs, documents, and connector data with retrieval and citation
Playbooks: Execute predefined multi-step workflows end to end
Agentic capabilities
Sub-agents: Delegate work to spawned sub-agents and merge results back into the primary workflow
MCP tool use: Discover and call external tools through MCP connectors
Content generation: Draft and edit long- and short-form content to a brief
Brand voice: Apply brand voice and custom instructions to output
Long-horizon execution: Sustain multi-hour objectives with planning, self-testing, and correction loops
Image analysis and generation: Interpret input images and generate new images
Presentations(early): Generate slide presentations from a prompt or source
Writer evaluates Palmyra X6 on production customer tasks (marketing, revenue, research, and outreach), not abstract reasoning benchmarks alone. Each capability is scored 0 to 1 against a locked baseline (n=22, reference frozen 2026-06-07). Scores measure Palmyra X6 running in WRITER Agent.Capability scores (Palmyra X6 in WRITER Agent)
Capability
What it measures
Score
Sub-agents
Delegating work to spawned sub-agents and merging results
0.86
Grounding and retrieval
Knowledge Graph and document grounding, retrieval, and citation
0.91
MCP tool use
Discovering and calling external tools through MCP connectors
0.88
Content generation
Drafting and editing long- and short-form content to a brief
0.88
Playbooks
Executing predefined multi-step workflows end to end
0.84
Model and system awareness
Knowing its identity, tools, scope, and when to safely refuse
0.97
Brand voice
Applying brand voice and custom instructions to output
0.92
Image analysis and generation
Interpreting input images and generating new images
0.80
Presentations (early)
Generating slide presentations from a prompt or source
0.74
Average capability score: 0.87.Domain benchmarks (raw model)The scores below measure the Palmyra X6 model directly, not in WRITER Agent, on finance benchmarks:
Benchmark
Score
Finance Agent
0.753
FinanceQA
0.673
FinanceBench
0.855
Palmyra X5
Deprecation notice: Palmyra X4 (palmyra-x4) and Palmyra X5 (palmyra-x5) will be deprecated on December 14, 2026.Migration path: Use Palmyra X6 (palmyra-x6) instead. Palmyra X6 has a 1M token context window (Palmyra X4 is 128k). See pricing for cost details and the deprecation policy for more information.
Overview
Use cases and capabilities
Benchmarking
Palmyra X5 is Writer’s general-purpose model for building and scaling AI agents, featuring a 1 million token context window, adaptive reasoning, and strong speed and cost efficiency.Pricing:
Input: $0.60 per 1M tokens
Output: $6.00 per 1M tokens
Palmyra X5 in WRITER Agent is billed at $5.00 input and $12.00 output per 1M tokens. See WRITER Agent pricing.
Content window: 1M
Palmyra X5’s 1M token context window further streamlines enterprise workflows and unlocks complex, multi-step use cases that weren’t possible before.Multi-step agentic workflows
Support documentation: Classify requests, assess urgency, assign a human review, stage updates in a CMS, and publish after approval.
Fund reporting: Streamline the analysis and preparation of detailed reports on the performance and status of investment funds, using reporting and research data pulled in from third-party systems
Content lifecycle management: Flag content that could be outdated, generate suggested revisions, and share them for human review.
Large data requirements:
Customer feedback analysis: Analyze large volumes of customer feedback to identify common themes, summarize sentiments, and generate actionable insights.
Research and development: Process and summarize multiple technical reports, research papers, and experimental data, accelerating innovation and product development.
Financial reporting: Process and summarize annual reports, SEC filings, and market analysis reports together at once to extract financial data, identify trends, and generate executive summaries.
Legal document analysis: Analyze lengthy legal documents, including contracts, patents, and compliance reports, to identify key clauses, flag potential risks, and ensure regulatory compliance.
Medical records analysis: Analyze and summarize large files of medical records containing structured and unstructured data, including patient records, clinical trial reports, audit reports, and more.
Palmyra X5 demonstrates robust performance across a suite of industry-standard benchmarks, showcasing its capabilities in reasoning, retrieval, and domain-specific tasks.
BBH (Big-Bench Hard): Evaluates complex reasoning and compositional logic. Palmyra X5 achieves a competitive score of 70.99%, aligning closely with top-tier models.
GPQA (Graduate-Level Google-Proof Q&A): Assesses the model’s ability to answer challenging, graduate-level questions in biology, physics, and chemistry that are resistant to simple lookup strategies. X5’s score of 47.20% indicates strong performance in scientific reasoning tasks.
MMLU_PRO: Focuses on professional-level knowledge across various domains such as law, medicine, and finance. Palmyra X5 scores 65.02%, demonstrating its suitability for enterprise applications in regulated sectors.
MATH_HARD: Tests symbolic reasoning and multi-step problem-solving abilities. X5’s score of 71.57% showcases its proficiency in handling complex analytical tasks.
Palmyra X4
Deprecation notice: Palmyra X4 (palmyra-x4) and Palmyra X5 (palmyra-x5) will be deprecated on December 14, 2026.Migration path: Use Palmyra X6 (palmyra-x6) instead. Palmyra X6 has a 1M token context window (Palmyra X4 is 128k). See pricing for cost details and the deprecation policy for more information.
Overview
Use cases and capabilities
Benchmarking
palmyra-x4 is an advanced language model that excels in processing and understanding complex tasks. It’s ideal for workflow automation, coding tasks, and data analysis.Pricing:
Input: $2.50 per 1M tokens
Output: $10.00 per 1M tokens
Content window: 128k
Agents & actions: Palmyra X4 acts as an advanced AI agent, capable of executing tasks beyond simple text generation by interacting with external systems like databases, applications, and other services. This enables it to perform real-time data updates and automate complex workflows.
Retrieval-augmented generation (RAG): Equipped with RAG, Palmyra X4 can retrieve and incorporate relevant information from vast data sources, enhancing the model’s accuracy and ensuring responses are always grounded in current, context-specific data.
Code generation: The model supports advanced code generation, enabling it to automate scripting and integrate seamlessly with various programming environments, optimizing workflows for technical teams.
Tool calling: Palmyra X4 is built to handle precise API interactions, allowing it to execute complex functions directly, making it a versatile tool for enterprise-level integrations and automated actions.
Palmyra X4 consistently ranks at the top in structured output, API tool calling, and accuracy for complex, multi-step workflows.
Top accuracy (ACC): Palmyra X4 achieves 78.76% accuracy in tool call identification and execution, leading the industry by nearly 20%.
Structured call planning (AST): Palmyra X4 scores 87.93% in planning and organizing tool calls, accurately interpreting input, generating parameters, and sequencing steps.
Execution performance (Exec): With an 88.27% score in executing tool calls, Palmyra X4 ranks highest in efficiently carrying out enterprise actions.
Global benchmarks: Palmyra X4 ranks in the world’s top 10 on HELM Lite (86.1%) and HELM MMLU (81.3%), excelling across 57 subjects.
To select the Palmyra model you want to use for a chat completion, specify the model ID in the model parameter of the request.Below is an example of specifying the palmyra-x5 model in a chat completion request.
from writerai import Writer# Initialize the client. If you don't pass the `api_key` parameter,# the client looks for the `WRITER_API_KEY` environment variable.client = Writer()chat_response = client.chat.chat( model="palmyra-x5", messages=[ { "role": "user", "content": "Summarize GDPR compliance requirements for a cloud-based data storage provider" } ])print(chat_response.choices[0].message.content)
import { Writer } from "writer-sdk";// Initialize the client. If you don't pass the `apiKey` parameter,// the client looks for the `WRITER_API_KEY` environment variable.const client = new Writer();const chat_response = await client.chat.chat({ model: "palmyra-x5", messages: [ { role: "user", content: "Summarize GDPR compliance requirements for a cloud-based data storage provider", }, ],});console.log(chat_response.choices[0].message.content);
You can also use prebuilt tools within a general-purpose chat completion request:
To select the Palmyra model you want to use for text generation, specify the model ID in the model parameter of the request.Below is an example of specifying the palmyra-x5 model in a text generation request.
curl --location 'https://api.writer.com/v1/completions' \--header 'Content-Type: application/json' \--header "Authorization: Bearer $WRITER_API_KEY" \--data '{ "model": "palmyra-x5", "prompt": "Summarize GDPR compliance requirements for a cloud-based data storage provider"}'
from writerai import Writer# Initialize the client. If you don't pass the `api_key` parameter,# the client looks for the `WRITER_API_KEY` environment variable.client = Writer()text_generation = client.completions.create( model="palmyra-x5", prompt="Summarize GDPR compliance requirements for a cloud-based data storage provider")print(text_generation.choices[0].text)
import { Writer } from "writer-sdk";// Initialize the client. If you don't pass the `apiKey` parameter,// the client looks for the `WRITER_API_KEY` environment variable.const client = new Writer();const text_generation = await client.completions.create({ model: "palmyra-x5", prompt: "Summarize GDPR compliance requirements for a cloud-based data storage provider",});console.log(text_generation.choices[0].text);
Get started with the Writer API by signing up for a free account and following the API quickstart.
We’ll announce the deprecation of a model at least three months in advance. This will give customers time to plan for the migration to the new model.
We’ll continue to support deprecated models for a period of time after they’re deprecated. This will give customers time to migrate to the new model.
We’ll eventually stop supporting deprecated models. The timeline for this will vary depending on the model. We will announce the end of support for a deprecated model at least six months in advance.