Arcee AI: Arcee Blitz

Text input Text output Unavailable
Author's Description

Arcee Blitz is a 24 B‑parameter dense model distilled from DeepSeek and built on Mistral architecture for "everyday" chat. The distillation‑plus‑refinement pipeline trims compute while keeping DeepSeek‑style reasoning, so Blitz punches above its weight on MMLU, GSM‑8K and BBH compared with other mid‑size open models. With a default 128 k context window and competitive throughput, it serves as a cost‑efficient workhorse for summarization, brainstorming and light code help. Internally, Arcee uses Blitz as the default writer in Conductor pipelines when the heavier Virtuoso line is not required. Users therefore get near‑70 B quality at ~⅓ the latency and price.

Key Specifications
Cost
$$$
Context
32K
Parameters
24B (Rumoured)
Released
May 05, 2025
Speed
Ability
Reliability
Supported Parameters

This model supports the following parameters:

Stop Presence Penalty Logit Bias Top P Temperature Min P Response Format Frequency Penalty Max Tokens
Features

This model supports the following features:

Response Format
Performance Summary

Arcee Blitz, a 24B-parameter dense model distilled from DeepSeek and built on Mistral architecture, demonstrates strong overall performance as a cost-efficient workhorse for various AI tasks. It consistently performs among the fastest models, ranking in the 79th percentile for speed across six benchmarks, and offers competitive pricing, placing in the 52nd percentile. A standout feature is its exceptional reliability, achieving a perfect 100th percentile, indicating minimal technical failures and consistent response delivery. In terms of specific benchmarks, Arcee Blitz exhibits notable strengths in General Knowledge (97.5% accuracy) and Ethics (99.0% accuracy), showcasing robust understanding and adherence to ethical principles. Its Coding performance is solid at 83.0% accuracy, though its duration is in the 91st percentile, suggesting it can be slower for coding tasks. For Classification, specifically Email Classification, it achieves high accuracy (95.0%) and is remarkably fast, being the most accurate among models of comparable speed. While its Reasoning (60.0% accuracy) and Instruction Following (59.2% accuracy) scores are moderate, they are competitive for a model of its size. The model's design for "everyday" chat and its internal use by Arcee for Conductor pipelines underscore its practical utility, offering near-70B quality at significantly reduced latency and price.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $0.45
Completion $0.75

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
Together
Together | arcee-ai/arcee-blitz 32K $0.45 / 1M tokens $0.75 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Free Executions Accuracy Cost Duration
Other Models by arcee-ai