AI tool comparison
Mistral Medium 3.5 vs Ternary Bonsai
Which one should you ship with? Here is the side-by-side panel verdict, pricing read, reviewer split, and community vote comparison.
AI Models
Mistral Medium 3.5
128B open-weight model with async remote coding agents and 256k context
75%
Panel ship
—
Community
Paid
Entry
Mistral Medium 3.5 is a 128B dense model with a 256k context window, scoring 77.6% on SWE-Bench Verified and 91.4 on τ³-Telecom. It's released with open weights under a modified MIT license — one of the strongest coding-capable open-weight releases this year. Priced at $1.50/M input and $7.50/M output via API, it's positioned as a cost-competitive alternative to proprietary frontier models for agentic and software engineering tasks. Alongside the model, Mistral is launching Vibe — a remote coding agent system that runs sessions in the cloud. Developers can start a task from the CLI or Le Chat, "teleport" their local session to the cloud (preserving history and approval state), and let it run asynchronously while they work on something else. Sessions run in isolated sandboxes and can automatically open pull requests on GitHub when complete. This competes directly with Devin, GitHub Copilot Workspace, and similar async coding agents. The Le Chat Work Mode adds a general-purpose agentic layer on top: multi-step workflows across email, calendar, and messaging, research synthesis from internal and external sources, and inbox triage with drafted replies. All actions are transparent and require explicit approval before anything sensitive executes. The combination of open weights, competitive pricing, and production-ready remote agents makes this one of Mistral's most significant releases since Mixtral.
Open Source Models
Ternary Bonsai
1.58-bit LLMs that run at 82 tok/s on M4 Pro and on your iPhone
75%
Panel ship
—
Community
Free
Entry
PrismML's Ternary Bonsai is a family of aggressively quantized language models that take the BitNet concept to its logical extreme. Each weight is constrained to one of three values — {-1, 0, +1} — with a shared FP16 scale factor per 128-weight group. No higher-precision escape hatches, no hybrid layers. The result is a 9x reduction in memory footprint versus standard 16-bit models. The numbers are striking: the 8B model fits in 1.75 GB and hits 82 tokens per second on an M4 Pro. More impressively, it runs at 27 tokens per second on an iPhone 17 Pro Max — fast enough for real-time conversation on-device. The 8B variant scores 75.5 average across standard benchmarks, outperforming many models that are 9-10x larger. The 4B and 1.7B variants push further into mobile-optimized territory. All three models are released under the Apache 2.0 license, available on Hugging Face and GitHub, and integrated into the Locally AI iOS app for immediate on-device deployment. For developers building privacy-sensitive applications or anyone tired of paying cloud inference costs, Ternary Bonsai offers a compelling on-device alternative that doesn't require a beefy GPU.
Reviewer scorecard
“Open weights at 77.6% SWE-Bench with cloud-native async agents is a compelling combo. The 'teleport local session to cloud' UX for Vibe is genuinely clever — it solves the context-loss problem when shifting from local to remote execution.”
“82 tokens per second on M4 Pro in 1.75 GB is a genuinely impressive engineering achievement. For local tooling, code assistants, or any latency-sensitive workload where I don't want cloud round-trips, this hits a sweet spot that larger quantized models miss. Apache 2.0 means I can embed it in commercial apps without legal headaches.”
“77.6% on SWE-Bench is strong but still behind Claude Sonnet and GPT-5.5 on the same benchmark. The Vibe agent is in 'public preview' which typically means rough edges. Wait for v1.0 before betting a production workflow on it.”
“A 75.5 benchmark average sounds good until you compare it against 8B models quantized with GGUF Q8 — which score similarly and have years of tooling, community support, and production deployments behind them. The 9x memory savings matter on constrained devices but less so on any machine with 16GB+ RAM. Niche but real use case.”
“Open-weight models with integrated remote agent infrastructure is the architecture that democratizes agentic AI. Any developer can self-host the weights and build their own agent backend — no vendor lock-in required.”
“On-device AI at 27 tokens per second on a phone is the inflection point that makes LLMs a platform primitive rather than a cloud service. Once inference is this cheap and fast on commodity hardware, the entire economic model of AI-as-API-call collapses. Ternary quantization is an early signal of where efficiency research is heading.”
“The Le Chat Work Mode covering email, calendar, and research synthesis is exactly what knowledge workers need. Mistral's approval-first approach to sensitive actions is the right balance between automation and human oversight.”
“The prospect of running a capable LLM entirely on my iPhone without sending any data to a server is genuinely exciting for creative work with sensitive material. Drafting, editing, and ideation without a cloud subscription or privacy concerns — I'd pay for that, and here it's free.”
Weekly AI Tool Verdicts
Get the next comparison in your inbox
New AI tools ship daily. We compare them before you waste an afternoon.