AI Tool Profile
MiniMax M3
MiniMax released M3 for coding, agent workflows, long-context reasoning, and native image and video understanding through MiniMax Code, Token Plans, and API access.
Verification & Sources
- Evidence state
- Recheck due
- Source links
- 8
- Freshness
- Needs recheck: checked August 4, 2026
- Last updated
- September 21, 2026
What this evidence state means
- Definition
- The claim was previously checked, but its review window expired or a material change may have invalidated it.
- Required provenance
- The prior evidence and check date are retained, together with the expiry or change signal that triggered recheck.
- Owner
- Kingy freshness queue owner and assigned editorial reviewer
- Freshness rule
- This is already outside its freshness rule. It must not be presented as current until reviewed against current evidence.
- Disputes and corrections
- Use “Suggest a correction” on the record. Kingy editorial reviews the cited evidence, records material corrections, and changes or removes the state when it is not supported.
Suggest a correction
MiniMax M3 is an open-weight, natively multimodal model from MiniMax for coding, agent workflows, long-context reasoning, and image or video understanding. MiniMax officially released M3 on June 1, 2026.
What MiniMax M3 does
MiniMax says M3 uses its MiniMax Sparse Attention architecture and supports context windows of up to one million tokens. The model is designed for software development, tool use, multi-step agent work, and multimodal inputs rather than text-only chat.
The official release describes two operating modes: a thinking mode for more complex reasoning and agent tasks, and a faster non-thinking mode for latency-sensitive work such as conversation or code completion. Those capabilities make M3 relevant to teams evaluating coding assistants, research agents, long-document workflows, and applications that need to combine text with images or video.
Access and pricing
MiniMax provides M3 through MiniMax Code, its Token Plans, and the MiniMax API. API pricing varies with input length and service tier, while Token Plans package model usage into monthly subscriptions. Buyers should confirm the current rates and limits on MiniMax’s official pages before estimating production cost.
Pricing reviewed September 21, 2026
The official pay-as-you-go guide still lists these M3 rates per million tokens, in input / output / cache-read order:
- Standard, up to 512K input tokens: $0.30 / $1.20 / $0.06.
- Standard, above 512K input tokens: $0.60 / $2.40 / $0.12.
- Priority, up to 512K input tokens: $0.45 / $1.80 / $0.09.
- Priority, above 512K input tokens: $0.90 / $3.60 / $0.18.
MiniMax labels these displayed rates a permanent 50% discount; the discount is already included. Priority costs 1.5 times Standard and is selected with service_tier: "priority". The threshold is input length. Standard Open Platform API keys draw from the pay-as-you-go account balance.
The current Token Plan documentation and signed-out Monthly purchase cards now show Plus $22/month, Max $55/month and Ultra $132/month. The Yearly cards displayed $220, $550 and $1,320 respectively, billed annually. MiniMax’s comparison table on the same page still shows $20/$50/$120 monthly. Kingy uses the current documentation and purchase-card amounts here; checkout and existing-subscriber renewal prices were not tested.
The cards describe approximately 1.7B, 5.1B and 12.5B M3 tokens per month. These are usage estimates: included quota has five-hour rolling and weekly limits, and unused quota does not carry into the next billing cycle. Text, image and speech share eligible quota. Music has been unavailable through Token Plan since August 20, 2026.
Token Plan uses a separate Subscription Key. Purchased Credits cost $1 per 1,000 credits, expire 365 days after purchase and can cover eligible overflow after included quota. Credits usage is charged at the resource’s pay-as-you-go list price; do not treat the advertised token estimates as an unrestricted allowance.
Amounts retain MiniMax’s $ notation. This is a source and signed-out browser review, not a paid usage or invoice test.
What to verify
Performance figures in MiniMax’s launch material are vendor-reported. Kingy.ai has not independently reproduced those benchmarks. Teams should test task reliability, latency, tool-call behavior, data handling, and total cost on their own workloads before moving M3 into production.
Official sources
- MiniMax M3 release announcement
- MiniMax M3 model overview
- MiniMax API documentation
- MiniMax pay-as-you-go pricing; checked September 21, 2026
- MiniMax Token Plan pricing; checked September 21, 2026
- MiniMax Token Plan purchase cards; browser checked September 21, 2026
The Kingy Brief
Get The Kingy Brief.
AI changes, original tests and one practical thing to try. Fridays at 09:00 Vancouver time.
Free · Double opt-in · Unsubscribe anytime
Sign up for The Kingy Brief on Beehiiv.
Signup help and newsletter schedule
Signup form provided by Beehiiv. After submitting, check your inbox for "Confirm your subscription to The Kingy Brief" and open its confirmation link. Check Spam or Promotions if you cannot find it.
Fridays at 09:00 Vancouver time: source-checked AI changes, original tests and one practical thing to try. The weekly restart begins October 9, 2026. We skip a week when there is not enough verified material. Free. Unsubscribe anytime.
Main competitors
Frontier coding and agentic models: Claude Opus 5 and Claude Sonnet 5 (Anthropic), the GPT-5.6 family (OpenAI) and Gemini 3.1 Pro (Google). Among open-weight releases the closest structural comparisons are DeepSeek V4 and Alibaba's Qwen3 family, which compete on the same open-weight, long-context, coding-first positioning.
Tool Links
Related Kingy Links
Launch History
MiniMax M3
MiniMax released M3 for coding, agent workflows, long-context reasoning, and native image and video understanding through MiniMax Code, Token Plans, and API access.
- Launch readiness
- 7.7 / 10
- Demo evidence
- Not scored yet
- Creator-story fit
- Not scored yet
Score definitions and rubric
These are launch-record readiness heuristics, not product ratings.
Launch readiness
How complete and reviewable the launch record is, not the quality of the product.
Inputs and weights: Launch date 15%; qualifying source 10%; what launched 10%; demo 15%; category 10%; audience 10%; editorial assessment 10%; traction evidence 10%; creator or audience fit 10%.
Evidence inputs: Reviewed launch metadata, public source links, demo links, taxonomy, audience, editorial notes, and recorded traction signals.
Demo evidence
Whether the record contains useful, reviewable demonstration evidence; it is not a rating of product output quality.
Inputs and weights: Working demo URL 45%; video walkthrough 25%; clear description of what launched 10%; audience 10%; editorial assessment 10%.
Evidence inputs: Demo and video URLs plus the reviewed launch description, audience, and editorial notes.
Creator-story fit
Whether a launch has enough demonstrable evidence and audience relevance for a useful creator story; it does not predict views or guarantee coverage.
Inputs and weights: Demo evidence 25%; visual creator category 15%; audience 15%; editorial assessment 15%; traction evidence 10%; pricing clarity 10%; API or open-weight evidence 10%.
Evidence inputs: Reviewed demo, category, audience, editorial, traction, pricing, API, and open-weight fields.
- Scale
- 0.0–10.0. A present qualifying input receives its published weight; a missing input receives zero. Scores are rounded to one decimal.
- Assigned by
- Suggested by the deterministic field-completeness helper and assigned or approved by a Kingy editorial reviewer.
- Rubric and check date
- Rubric version P0-2026-08-10. The record’s “Last verified” date is the score check date. Checked: 2026-07-09.
- Confidence and missing data
- Confidence depends on source completeness. “Not scored yet” means no reviewed value; “Needs review” means the value or score set failed validation.
- Freshness
- Recalculate after a material launch, source, demo, pricing, audience, or traction change and during the record freshness review.
- Disputes
- Use “Suggest a correction” on the record and cite the relevant evidence. Commercial relationships cannot buy or alter a score.
MiniMax released M3, an open-weight model for coding, agent workflows, long-context reasoning, and native image and video understanding, available through MiniMax Code, Token Plans,…