WizardLM
Microsoft's instruction-following LLM trained with Evol-Instruct — progressively harder instructions produce top results.
Overview
WizardLM is a series of open-source instruction-tuned language models from Microsoft Research that applies the Evol-Instruct training methodology to general instruction following. The same technique used for WizardCoder (code specialisation) is applied here to general instruction data, progressively evolving simple instructions into more complex, nuanced versions during training.
WizardLM-70B achieved performance competitive with GPT-3.5 on instruction-following benchmarks while WizardLM-13B was competitive with models twice its size. The Evol-Instruct methodology generates instruction variants that are progressively harder: adding constraints, requiring multi-step reasoning, combining tasks, and introducing abstract complexity. This exposes the model to a much wider range of instruction difficulty during training.
WizardLM models are widely used for research into instruction following quality and as capable open models for deployment. The family includes WizardMath (mathematical reasoning) and WizardCoder (code generation) as domain-specific variants, making WizardLM a comprehensive family for instruction-following applications.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 8 May 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
Use cases
ASEAN Perspective
WizardLM in Southeast Asia
ASEAN-region availability and pricing notes coming soon. Drop the editorial team a note via /contact/ if you can supply local context (Singapore/Malaysia/Indonesia/Thailand/Vietnam).
WizardLM is an open-source large language model family that popularised the Evol-Instruct method of automatically evolving training instructions to richer complexity, producing instruction-following models that punched above their weight at launch. As open weights with no usage cost, it is valuable for research, fine-tuning experiments and on-prem deployments where data cannot leave the building.
The wider open-LLM field has moved well past these models, and project activity has slowed, so it is now more of a historical/reference choice than a leading chat model. There is no hosted service or API, only weights you must run yourself, which requires GPUs and ML skills. Appropriate for ML practitioners, not for general users wanting a ready-made chatbot.
Notable facts
- WizardLM's Evol-Instruct creates instructions that are 4x more complex on average than the original training data, teaching models to handle requests that simple instruction datasets never include.
- The WizardLM family (WizardLM, WizardCoder, WizardMath) was produced by a team of 3 Microsoft Research interns in under 6 months.
- WizardLM-70B was the first open model to achieve GPT-3.5-level performance on the MT-Bench conversational evaluation.
Frequently asked questions
About this listing
This entry was compiled from publicly available data including WizardLM's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with WizardLM unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to WizardLM directly →
Spotted something out of date? Suggest an update →
Alternatives to WizardLM
More in LLMs & Chat
