Changelog

1.9.6 (2026-04-14)

Multiview Image Workflow Expansion

We've expanded multiview image support to improve 3D asset generation pipelines and reduce manual preprocessing.

New task types

• Added generate_multiview_image and edit_multiview_image, enabling generation of multi-view image sets of the same subject from a single input image, suitable for 3D reconstruction workflows (e.g. model generation).
• Each generated view can be edited independently, enabling targeted refinements without regenerating the full multiview set.
• Dedicated endpoint documentation available at: generate-multiview-image.md

Improved pipeline integration

• multiview_to_model now accepts an optional parameter: original_task_id.
• This allows direct chaining from multiview generation/edit tasks without re-uploading image files.
• Constraint:
  • files and original_task_id are mutually exclusive.
  • original_task_id only supports outputs from:
    • generate_multiview_image
    • edit_multiview_image
• This simplifies pipelines where multiview images are generated programmatically.

Documentation updates

• Standardized multiview task output schema under: task.md.
• Updated billing documentation for the new task types.
• Updated schema definitions for consistency across multiview endpoints.

1.9.5 (2026-03-11)

New Models Available

• Tripo v3.1: Released Tripo v3.1 for high-fidelity 3D generation, supporting up to 2M polygons for 3D printing, hero assets, and high-detail visuals.

• Tripo P1.0: Released P1-20260311, a new Smart Mesh model for structured mesh generation, optimized for clean topology, ~2s mesh generation, and real-time production workflows. P1-20260311 supports only selected generation parameters. Unsupported parameters will return an error. See the parameter documentation for details.

• Tripo Rig v2.5

1.9.4 (2026-03-09)

• Large Image Handling Fix: Fixed an issue where generation could fail when the input image resolution or size was too large.

• Autofix Optimization: Improved the autofix parameter in the model generation API, enabling better handling of common image issues, including blurry inputs, incomplete image information, and similar cases.

• Credit Usage in Task Response: Successful responses for credit-consuming tasks now include credit usage for that specific request, making it easier for developers to track and manage usage.

1.9.3 （2025-12-30）

• Performance Enhancement: Significantly accelerated generation speed for Tripo v3.0 base geometry models, reducing processing time for untextured mesh output.
• Texture Quality Improvement: Optimized the texturing pipeline for Tripo v3.0, delivering enhanced color accuracy, improved surface detail fidelity, and more refined material representation.
• Low-Poly Model Addition: Introduced Tripo v2.0 low-polygon model generation with enhanced quality, offering improved geometric topology and better preservation of key features at reduced polygon counts.

1.9.2 （2025-11-04）

• Multiview-to-model is now supported with Tripo v3.0 (Standard & Ultra)

1.9.1 (2025-09-26)

• Tripo v3.0 Official Launch:
  • The Tripo v3.0 model introduces significant enhancements in geometric interpretation and output. Assets now generate with crisper edges, cleaner surfaces, and an unparalleled level of structural coherence.
  • Ultra Mode unlocks the ability to generate assets with up to 2 million polygons, capturing maximum detail and fidelity for the most demanding creations. Set geometry_quality to detailed to try.
  • The texturing and material generation pipeline has been substantially upgraded for more accurate and physically-based results.
  • Clarity and precision for fine details, including embedded text and complex patterns on surfaces, have been greatly improved.
  • Materials now exhibit more accurate light interaction, with well-balanced colors and natural surface finishes suitable for modern PBR pipelines.
• Improvements Since Beta
  • The geometry generation model has been refined for improved structural soundness. This results in more coherent details and superior handling of complex forms.
  • The geometry generation process has been optimized to resolve critical issues, including the elimination of double-layer surface artifacts. We have also improved surface continuity to significantly reduce the occurrence of gaps or holes
  • Better texture–image alignment, ensuring higher color fidelity

1.9.0 (2025-08-12)

• Added:
  • New version v3.0-20250812 (BETA). The detail restoration is greatly improved, the edges are sharper, the hard surface support is better
  • Texture Gen Updated to v3.0-20250812 (BETA). Clarity and realism are both improved, and text rendering capabilities are enhanced.
  • PBR Generator Enhanced. Present more realistic material textures and make the model more tangible

1.8.0 (2025-06-18)

This release focuses on migrating Studio's core algorithm capabilities to the OpenAPI and introducing significant feature enhancements.

• Added:

  • Introduced new mesh editing task types: mesh_segmentation, smart_low_poly, and mesh_completion. Refer to Mesh Editing for more details.
  • Added text_to_image task type for generating images from text prompts.
  • Enhanced text_to_model, image_to_model, and multiview_to_model tasks with smart_low_poly and generate_parts parameters for finer control over model generation.
  • The convert task now supports additional parameters for advanced customization.
  • All generation task types now support the compress parameter for output optimization.
  • The texture_model task now includes a prompt parameter, accepting text prompt, image prompt, and style_image as style reference for precise control over texture generation.
  • Added the animations parameter to the retarget task, allowing multiple animations in a single request.
  • The pre_rig_check task now returns a rig_type
  • The animate_rig task now accepts parameter spec (options: mixamo or tripo) and rig_type to specify the rigging method

• Modified:

  • Refactored the API schema documentation (schema.md) for improved clarity and organization.

1.6.0 (2025-05-21)

• Added: New model Turbo-v1.0-20250506 for text-to-model and image-to-model generation featuring significantly faster processing speed, optimized for time-sensitive applications.
• Modified: Refactored schema.md for improved clarity and organization.

1.5.0 (2025-01-23)

• Added: New version v2.5-20250123 featuring enhanced geometry details and realistic textures.
• Added: Two new styles: gold and ancient_bronze.
• Added: quad parameter for generation interfaces.
• Modified: Default version changed from v1.4-20240625 to v2.5-20250123.
• Deprecated: Version v1.3-20240522 and multiview generation of version 1.4-20240522 are now deprecated.

1.4.9 (2024-12-06)

• Added: orientation parameter for image_to_model and multiview_to_model.
• Added: Additional preset animations for retargeting.
• Added: Two new styles and style improvements.

1.4.8 (2024-11-20)

• Added: style parameter for text_to_model generation.
• Added: auto_size parameter for text_to_model, image_to_model, and multiview_to_model.
• Modified: Conversion format can now be performed based on a previously converted model.
• Modified: face_limit for conversion can now be applied without requiring quad.

1.4.7 (2024-11-12)

• Added: url support for image_to_model and multiview_to_model generation.

1.4.6 (2024-11-11)

• Added: style parameter for image_to_model generation.
• Added: v2.0-20240919 support for multiview_to_model generation.
• Added: New texture_model endpoint.
• Added: texture_alignment for image_to_model, multiview_to_model, and texture_model.
• Added: texture_quality for text_to_model, image_to_model, multiview_to_model, and texture_model.

1.4.5 (2024-10-12)

• Added: Support for 3mf conversion format.

1.4.4 (2024-09-19)

• Added: Completely new algorithm version (v2.0-20240919).
• Added: Ability to watch all tasks since the last check.
• Added: scale_factor parameter for convert.

1.4.2 (2024-09-02)

• Added: Multiview to Model method.
• Added: Check Balance endpoint.
• Added: New parameters for post-processing.

1.1.2 (2024-07-30)

• Added: Streaming methods for retrieving models.
• Modified: Replaced model with model_version for specifying model versions.
• Modified: Simplified error codes.

1.1.1 (2024-07-09)

• Added: New parameters for various endpoints.
• Added: Model version support for draft generation.

1.1.0 (2024-05-06)

• Added: New parameters for various endpoints.
• Modified: Split the animation API into three separate endpoints.
• Modified: Removed callbacks for stylizing and conversion; now using polling instead.

1.0.0 (2024-04-08)

• Added: The first API billing plan, including pricing and FAQ.

0.2.0-beta (2024-02-27)

• Added: New API version and endpoint changes.
• Added: Improved documentation.

0.1.1-beta (2024-01-09)

• Added: Initial release of the Image to 3D API.
• Modified: Enhancements to expression parsing and API schema.

0.1.0-beta (2024-01-03)

• Initial Release: The first version of our API.