Friendify’s latest update adds Discord image generation, a redesigned bot management experience, smarter model selection, premium model controls, safer provider-filter handling, and consent-based analytics.
Friendify Team · July 11, 2026
Friendify has a new product update focused on Discord, model control, character editing, safety, and privacy.
This release builds on the previous big update, but the focus this time is more specific: giving Discord bots stronger creative tools, making bot management easier, improving how models are selected, and giving users clearer control over analytics and privacy.
Discord bots on Friendify can now create character images directly inside Discord.
This makes Discord characters feel more complete without needing users to leave the server. A character can now chat, roleplay, and generate visuals from the same Discord experience.
Discord image generation also includes bot-level controls, so creators can decide how image generation should work for each bot.
Discord image generation is not just a simple on/off feature.
Bot owners now have access to settings for usage limits, guest quotas, access rules, and image generation behavior. This makes it easier to control who can generate images, how often they can use the feature, and how the bot should handle image requests.
This is especially useful for public servers, community bots, and shared characters where unrestricted image generation could get messy fast.
The Discord bot management experience has been rebuilt to be clearer and easier to use.
The new flow includes a better bot creation wizard, a redesigned bot management window, channel and role access controls, quota management, and dedicated image generation settings.
Instead of hiding important controls across different places, the new management experience groups them into clearer sections. Managing a Discord bot should now feel much less like digging through a settings maze.
The slash command enable/disable setting is now enforced by the Discord gateway.
That means when slash commands are turned off for a bot, the runtime now respects that setting properly instead of only showing it in the interface.
Small detail, big sanity improvement.
Character and chat model selection has moved into a new modal experience.
This makes choosing a model easier to understand, especially as Friendify supports more model options across different plans and use cases.
There is also a new Automatic model option. Instead of manually choosing every time, users can let Friendify pick the best available option based on the current setup.
Premium models are now shown more clearly in the model list.
If a model is not available on the user’s current plan, it appears locked instead of disappearing completely. The interface also explains why the model cannot be used.
This should make upgrades and model limits easier to understand without making the model list feel confusing or broken.
The admin model management screen has been reorganized around access, plan availability, and usage policy.
Models can now have minimum subscription plan requirements, making it easier to decide which models belong to Free, Spark, Bond, or Realm-level access.
Image prompt composer models were also improved. Composer models can now be assigned separately from the models users directly see or choose.
Friendify can also use separate composer models for SFW and NSFW image prompts, with only one active composer model per content level. This keeps the image prompt pipeline cleaner and easier to manage.
The character creation and editing screens have been reorganized into clearer feature sections.
Marketplace publishing settings now have their own visible section inside the character editor, so creators can more easily understand how and when a character is published.
The onboarding tour has also been updated to explain the new character editor structure, helping new users understand where important settings live.
Sometimes a model provider refuses a response because of its own safety filter.
Those refusals are now handled more cleanly. Instead of saving the refusal as if it were a normal character message, Friendify can avoid adding it to the conversation and retry in a controlled way when appropriate.
This helps keep chats cleaner and prevents safety-filter artifacts from becoming part of the character’s conversation history.
Friendify now supports consent-based Google Analytics and Google Ads measurement.
Analytics and advertising tags do not run unless the user gives permission. Users can also change their analytics preferences later through the cookie settings.
The cookie consent screen, privacy page, and cookie page were updated to match this new measurement system.
This gives Friendify better product insight while keeping user choice clear.
Friendify links now have a more branded OpenGraph preview image when shared on social platforms.
That means links should look cleaner and more recognizable when posted on places like Discord, X, Reddit, or other social feeds.
This update makes Friendify stronger in three important areas:
Discord bots are more creative.
Model access is easier to understand.
Privacy and safety controls are cleaner.
It is not just a visual update. It improves how creators manage bots, how users choose models, how admins control model access, and how Friendify handles privacy-friendly measurement.