
Arcee AI
Notable people
People map
Current peopleNotable public-facing people we cover. Grouped by public roles, not reporting lines; this is not a complete staff directory.
Leadership
- Mark McQuade
Co-founder and chief executive officer
- Lucas Atkins
Chief technology officer and head of research
Other public roles
- Allison Durham
Member of Technical Staff
Former people
Role status unverified
- Fernando Fernandes Neto
Research contributor · Status unverified
Writing, research & talks
Arcee released five models as open weights: SuperNova-v1 70B, Virtuoso-Large 72B, the tool-calling Caller 32B, GLM-4-32B-Base-32K and Homunculus 12B. Three had been production-only on its SaaS, with licenses ranging from Llama 3 Community to Apache-2.0.
Spectrum trains only the layers with a high signal-to-noise ratio and freezes the rest. Arcee reports up to 42 percent faster training with quality held, lower memory use allowing larger batches, and less catastrophic forgetting because untouched layers keep prior knowledge.
Three lessons from published MergeKit use cases, led by Pre-trained Model Average: averaging checkpoints from the stable phase of pre-training improves performance, with merge interval and model count both mattering. Arcee used the method building AFM-4.5B.
What breaks when adapting an LLM to a domain: full-weight fine-tuning erodes general capability, real customer corpora are noisy, templated and narrow, and filling the gap with synthetic data brings its own grounding and alignment risks. Argues for smart filtering over volume.
MergeKit can decompose the difference between a fine-tune and its base model into a PEFT-compatible low-rank adapter. On OpenHermes 2.5, a rank-8 extraction holding 0.46 percent of the parameters kept a 9.5 percent average gain over base Mistral; rank 512 reached 90 percent of full performance.
- Arcee Meraj Maarjarcee.ai
Arcee-Meraj adapts the Qwen2-72B-based Nova model to Arabic through a curated bilingual data mixture, batch translation of English sets, task-specific cleaning, self-curation and iterative training with DPO. Arcee reports it leading the Open Arabic LLM Leaderboard.
- Virtuoso Lite Virtuoso Medium V2 Distilling Deepseek V3 Into 10b 32b Small Language Models Slmsarcee.ai
Arcee distilled DeepSeek-V3 into a 10B Falcon-based model and a 32B, both Apache-2.0. Virtuoso-Lite trained on the V3 tokenizer to absorb 1.1B tokens of logits from an fp8 V3, then reverted to a Llama-3 tokenizer and merged. The 32B outscores their earlier 72B Nova.
Arcee ran continual pre-training on Mistral-7B-Instruct over 2.5B tokens of SEC filings twice, once full and once with QLoRA. Perplexity landed at 2.72 for standard CPT against 3.30 for QLoRA and 3.96 for the base, evidence that low-rank updates cannot absorb substantial new knowledge.
A walkthrough of serving AFM-4.5B on Intel Xeon 6 using OpenVINO and Hugging Face Optimum Intel: quantize and convert in a single CLI command, then serve through OpenVINO Model Server behind an OpenAI-compatible endpoint, benchmarked on an r8i.8xlarge instance.
How Differentiable Adaptive Merging replaces evolutionary model merging with learned per-column scaling coefficients over weight matrices, with the researcher explaining why the optimisation framing is faster and more predictable than parameter search.
- Mergekit Returns To Its Rootsarcee.ai
Arcee returned Mergekit to LGPL v3 after its Business Source License period, citing adoption friction from legal review. Commercial and proprietary use is unrestricted; distributing modified copies of the library requires releasing those changes.
- The Trinity Manifestoarcee.ai
Arcee launched the Trinity family: Mini as a post-trained reasoning MoE and Nano Preview as an experimental high-sparsity chat model, with Trinity Large in training on 2048 B300 GPUs. The post explains why the team moved from post-training others' checkpoints to owning pretraining.
A walkthrough of converting AFM-4.5B into a hybrid Kimi Delta Attention and global attention model through knowledge distillation, covering student initialisation, dropping RoPE without replacement, and a modified RADLADS pipeline ending in long-context finetuning.
- Trinity Largearcee.ai
Arcee's 400B sparse MoE, trained on 2048 B300 GPUs in about 30 days, released as Preview, Base and a 10T-token TrueBase checkpoint. Covers the 4-of-256 routing choice, extra dense layers for routing stability, momentum-based expert balancing and z-loss.


