v1.5.9

v1.5.9 changelog - 2026-06-07

NPX

npx -y @maximhq/bifrost --transport-version v1.5.9

Docker

docker pull maximhq/bifrost:v1.5.9
docker run -p 8080:8080 maximhq/bifrost:v1.5.9

​

Bifrost(HTTP)

1.5.9

Private Network Access — To connect to a provider on a private network (e.g. a local vLLM or Ollama instance), set allow_private_network: true in the provider’s network_config. This allows connections to RFC 1918 ranges (192.168.x.x, 10.x.x.x, 172.16.x.x). Link-local addresses (169.254.x.x) remain blocked regardless of this setting. See Provider Configuration for details.

✨ Features

• OpenAI Compaction — Added OpenAI conversation compaction support across core, framework, logging, and the API surface (#4053)
• Multi-Customer & Org Hierarchy — Logs and usage tracking now support multiple customers, teams, and business units, including business unit CRUD, team assignment, and governance endpoints in the OpenAPI spec (#4066, #4041, #4082)
• Provider-Level Governance — Budgets & limits are now scope-aware and can be applied at the virtual-key top level and per provider, wired from the model configs table, with UI filters for scope and providers (#3938, #3937, #3939, #3981, #3962)
• Customer Budgets — Customers support multiple budgets and calendar_aligned budget windows (#3998, #3997)
• Virtual Key Attribution & Controls — Added a created_by user attribution column and a blacklisted_models column for virtual key provider configs (#3672, #3653)
• Request Header Capture — OTel and Maxim observability plugins capture request_headers by pattern, with wildcard support (e.g. x-custom-*); logging gained the same wildcard header capture (#4012, #3958)
• OTel Content Controls & Collectors — New disable_content_logging option drops message/tool content from exported spans, plus support for multiple OTel collectors (#4064, #3894)
• xAI x_search — Added xAI x_search tool support (#3976)
• URL Validation — Added fetch URL validation with private-network configuration and link-local blocking (#3947, #3991)
• File Scheme Pricing URLs — Pricing source URLs now accept the file:// scheme for air-gapped and self-hosted deployments (#4045)
• Paginated Virtual Keys — Virtual key fetching is paginated to handle deployments with very large numbers of keys (#3957)
• Client IP Resolution — Resolve client IP from X-Forwarded-For/X-Real-IP headers
• SCIM Provisioning — Added attributeType/attributeValue SCIM provisioning fields
• Helm/Config Schema — Added roles RBAC governance config and per_user_oauth MCP auth to the Helm chart and config schema (#4004, #4009)
• Log Navigation UI — Added a “View logs” menu item to customer, team, and virtual key tables, clickable links in log detail views, a customer detail sheet, and a reusable BudgetDisplay component (#4073, #4054, #4026, #4055)
• Faster First Paint — Added an inline loading shell to #root before React mounts (#4063)
• Materialized View Alias — Added an alias column to the materialized view with filter support (#4078)

🐞 Fixed

• Fetch URL IP Checks — Hardened fetch URL IP checks against SSRF (#4092)
• Mantle Model Matching — Broadened Mantle model matching to all gpt variants (#4091)
• Empty Thinking Blocks — Strip thinking blocks when the signature is empty (#4079)
• OpenAI Stream Usage — Removed usage from the responses.created event in the OpenAI stream (#4080)
• Prompt Cache Key — Set the prompt cache key from the Anthropic integration (#4086)
• Upstream Failure Status — Map upstream connection failures to 502 instead of 400 (#3929) (thanks @chris-colinsky!)
• Gemini Schema Constraints — Accept numeric schema integer constraints for Gemini (#3994) (thanks @yanhao98!)
• Files Provider Param — Accept the ?provider= query param on GET /v1/files (#3971) (thanks @alexef!)
• Optional Batch Model — Made the model field optional on POST /v1/batches (#3973) (thanks @alexef!)
• Helm Azure Config — Added missing azure_key_config fields to the Helm schema (#3996) (thanks @axelray-dev!)
• Text Completion Chunk Model — Added the missing Model field to TextCompletionChunkResponse (#3970) (thanks @kuishou68!)
• MCP Inline stdio Env — MCP stdio server configs accept inline environment variable assignments (#3861) (thanks @Shushmitaaaa!)
• Orphaned Tool Results — Orphaned tool results in the OpenAI to Anthropic conversion flow are no longer rejected by the Anthropic API (#3919)
• Node Usage Reconciliation — Added a monotonic inc_number log cursor so node usage reconciliation does not skip late async log writes (#3664)
• Bedrock Output Assessments — Corrected the type of outputAssessments in Bedrock responses (#4028)
• Model Pool Pricing Reloads — Preserve non-pricing model pool entries across pricing reloads (#3999)
• Ghost Node Reconciliation — Replicate the VK hierarchy flow for ghost node reconciliation (#4088)
• VK Double Usage Counting — Fixed double usage counting when creating a virtual key (#4070)
• Model Config Lifecycle — Cascade deletes for model configs and removal of stale in-memory model configs (#4051, #4043)
• FTS Index Cap — Reduced the FTS index left() cap from 800k to 250k chars to stay within the tsvector limit (#4057)
• Sync Worker Drift — Reduced the sync worker ticker period to 5m to prevent threshold drift (#4023)
• Passthrough — Fixed passthrough budgets, gated passthrough models per VK, model extraction for Azure passthrough, and restricted fallbacks/provider selection to the VK boundary (#3941, #3988, #3983, #3924)
• Provider Response Headers — Strip provider response headers and add a content-type filter (#3955, #4024)
• Stream Handling — Drain non-SSE stream readers and retry stale connections (#3956, #3967)
• Azure Claude — Strip Azure diagnostic property for Claude models (#3925)
• Compat max_tokens — Preserve chat max_tokens during param filtering (#3992)
• Raw Request Flag — Removed the raw request flag from providers that don’t support it (#4058)
• UI Fixes — Standardized page container layout, virtual key model configs UI, and dashboard chart tooltips (#4046, #4052, #4044)

🔧 Maintenance

• Dependency Upgrades — Bumped transitive golang.org/x dependencies (crypto, net, sys, text) for Docker Scout CVE remediation and recharts to 3.8.1; cascaded version bumps across all modules (#3900, #4003)

⏲️ Rolling back (to 1.5.8)

These are the DB queries you would need to fire if you need to rollback to v1.5.8.

Single node setup

-- ============================================================
-- STEP 1: Rollback migrate_virtual_key_governance_to_model_configs
-- ============================================================

BEGIN;

-- 1A. Restore VK top-level governance (provider IS NULL rows)

UPDATE governance_budgets
SET virtual_key_id = mc.scope_id,
    model_config_id = NULL
FROM governance_model_configs mc
WHERE governance_budgets.model_config_id = mc.id
  AND mc.scope = 'virtual_key'
  AND mc.model_name = '*'
  AND mc.provider IS NULL
  AND mc.scope_id IS NOT NULL;

UPDATE governance_virtual_keys
SET rate_limit_id = mc.rate_limit_id
FROM governance_model_configs mc
WHERE mc.scope = 'virtual_key'
  AND mc.scope_id = governance_virtual_keys.id
  AND mc.model_name = '*'
  AND mc.provider IS NULL
  AND mc.rate_limit_id IS NOT NULL;

-- 1B. Restore VK per-provider governance (provider IS NOT NULL rows)

UPDATE governance_budgets
SET provider_config_id = pc.id,
    model_config_id = NULL
FROM governance_model_configs mc
JOIN governance_virtual_key_provider_configs pc
  ON pc.virtual_key_id = mc.scope_id AND pc.provider = mc.provider
WHERE governance_budgets.model_config_id = mc.id
  AND mc.scope = 'virtual_key'
  AND mc.model_name = '*'
  AND mc.provider IS NOT NULL
  AND mc.scope_id IS NOT NULL;

UPDATE governance_virtual_key_provider_configs
SET rate_limit_id = mc.rate_limit_id
FROM governance_model_configs mc
WHERE mc.scope = 'virtual_key'
  AND mc.scope_id = governance_virtual_key_provider_configs.virtual_key_id
  AND mc.provider = governance_virtual_key_provider_configs.provider
  AND mc.model_name = '*'
  AND mc.rate_limit_id IS NOT NULL;

-- 1C. Delete the VK-scoped wildcard model config rows
DELETE FROM governance_model_configs
WHERE scope = 'virtual_key'
  AND model_name = '*';

-- 1D. Remove the migration record
DELETE FROM migrations WHERE id = 'migrate_virtual_key_governance_to_model_configs';

COMMIT;

-- ============================================================
-- STEP 2: Rollback migrate_provider_governance_to_model_configs
-- ============================================================

BEGIN;

UPDATE config_providers
SET budget_id = mc.budget_id,
    rate_limit_id = mc.rate_limit_id
FROM governance_model_configs mc
WHERE mc.scope = 'global'
  AND mc.scope_id IS NULL
  AND mc.model_name = '*'
  AND mc.provider = config_providers.name;

DELETE FROM governance_model_configs
WHERE scope = 'global'
  AND scope_id IS NULL
  AND model_name = '*'
  AND provider IS NOT NULL;

DELETE FROM migrations WHERE id = 'migrate_provider_governance_to_model_configs';

COMMIT;

-- ============================================================
-- STEP 3 (if needed): Rollback add_budget_model_config_id_column
-- Only if downgrading to a version before this column existed
-- ============================================================

BEGIN;

UPDATE governance_budgets SET model_config_id = NULL WHERE model_config_id IS NOT NULL;

ALTER TABLE governance_budgets DROP COLUMN model_config_id;

DELETE FROM migrations WHERE id = 'add_budget_model_config_id_column';

COMMIT;

Multiple nodes

-- ============================================================
-- PHASE 1: DUAL-WRITE — restore old FKs, keep new rows intact
-- Safe to run while the current (new) version is serving traffic
-- ============================================================

BEGIN;

-- 1A. Restore provider-level governance back to config_providers
UPDATE config_providers
SET budget_id = mc.budget_id,
    rate_limit_id = mc.rate_limit_id
FROM governance_model_configs mc
WHERE mc.scope = 'global'
  AND mc.scope_id IS NULL
  AND mc.model_name = '*'
  AND mc.provider = config_providers.name;

-- 1B. Restore VK top-level budgets back to governance_virtual_keys
--     (intentionally NOT clearing model_config_id yet — new pods still need it)
UPDATE governance_budgets
SET virtual_key_id = mc.scope_id
FROM governance_model_configs mc
WHERE governance_budgets.model_config_id = mc.id
  AND mc.scope = 'virtual_key'
  AND mc.model_name = '*'
  AND mc.provider IS NULL
  AND mc.scope_id IS NOT NULL;

-- Restore VK top-level rate limits
UPDATE governance_virtual_keys
SET rate_limit_id = mc.rate_limit_id
FROM governance_model_configs mc
WHERE mc.scope = 'virtual_key'
  AND mc.scope_id = governance_virtual_keys.id
  AND mc.model_name = '*'
  AND mc.provider IS NULL
  AND mc.rate_limit_id IS NOT NULL;

-- 1C. Restore VK per-provider budgets back to provider configs
UPDATE governance_budgets
SET provider_config_id = pc.id
FROM governance_model_configs mc
JOIN governance_virtual_key_provider_configs pc
  ON pc.virtual_key_id = mc.scope_id AND pc.provider = mc.provider
WHERE governance_budgets.model_config_id = mc.id
  AND mc.scope = 'virtual_key'
  AND mc.model_name = '*'
  AND mc.provider IS NOT NULL
  AND mc.scope_id IS NOT NULL;

-- Restore VK per-provider rate limits
UPDATE governance_virtual_key_provider_configs
SET rate_limit_id = mc.rate_limit_id
FROM governance_model_configs mc
WHERE mc.scope = 'virtual_key'
  AND mc.scope_id = governance_virtual_key_provider_configs.virtual_key_id
  AND mc.provider = governance_virtual_key_provider_configs.provider
  AND mc.model_name = '*'
  AND mc.rate_limit_id IS NOT NULL;

COMMIT;


-- ============================================================
-- PHASE 2: CLEANUP — remove new-version-only data
-- Run ONLY after ALL pods are on the old version
-- ============================================================

BEGIN;

-- Clear model_config_id from budgets (old version doesn't use it)
UPDATE governance_budgets SET model_config_id = NULL WHERE model_config_id IS NOT NULL;

-- Delete the VK-scoped wildcard model config rows
DELETE FROM governance_model_configs
WHERE scope = 'virtual_key'
  AND model_name = '*';

-- Delete the provider-level wildcard model config rows
DELETE FROM governance_model_configs
WHERE scope = 'global'
  AND scope_id IS NULL
  AND model_name = '*'
  AND provider IS NOT NULL;

-- Remove migration records so a future upgrade re-runs them
DELETE FROM migrations WHERE id = 'migrate_virtual_key_governance_to_model_configs';
DELETE FROM migrations WHERE id = 'migrate_provider_governance_to_model_configs';

-- Optional: only if the old version predates the model_config_id column
-- ALTER TABLE governance_budgets DROP COLUMN model_config_id;
-- DELETE FROM migrations WHERE id = 'add_budget_model_config_id_column';

COMMIT;

​

Core

1.5.17

• feat: OpenAI compaction support (#4053)
• feat: multiple customers logs and usage tracking (#4066)
• feat: multiple team and business unit support in logstore (#4041)
• feat: request_headers pattern capture for OTel and Maxim plugins with wildcard support (#4012)
• feat: xAI x_search tool support (#3976)
• feat: fetch URL validation with private-network config and link-local blocking (#3947, #3991)
• feat: file:// scheme support for pricing URLs (#4045)
• feat: filter ListAllModels provider fan-out by virtual key’s allowed providers (#3796)
• fix: harden fetch URL IP checks against SSRF (#4092)
• fix: broaden Mantle model matching to all gpt variants (#4091)
• fix: strip thinking block if signature is empty (#4079)
• fix: remove usage from responses.created event in OpenAI stream (#4080)
• fix: set prompt cache key from Anthropic integration (#4086)
• fix: map upstream connection failures to 502 instead of 400 (#3929) (thanks @chris-colinsky!)
• fix(gemini): accept numeric schema integer constraints (#3994) (thanks @yanhao98!)
• fix: handle compaction message type (#3966)
• fix: OpenAI integration content string handling (#3949)
• fix: capture resolved provider from the load balancer for logging (#3930)
• fix: add content type to provider response header filter (#4024)
• fix: strip provider response headers (#3955)
• fix: drain non-SSE stream reader (#3956)
• fix: stale connection retries (#3967)
• fix: Azure diagnostic property strip for Claude models (#3925)
• fix: passthrough budgets (#3941)
• fix: bedrock outputAssessments type correction (#4028)
• fix: add Model field to TextCompletionChunkResponse (#3970) (thanks @kuishou68!)
• fix: accept orphaned tool results in OpenAI to Anthropic conversion flow (#3919)
• fix(mcp): allow inline stdio env assignments (#3861) (thanks @Shushmitaaaa!)
• chore: bumped transitive golang.org/x dependencies (crypto, net, sys, text) for Docker Scout CVE remediation (#3900)

​

Framework

1.3.17

• feat: multiple customers logs and usage tracking (#4066)
• feat: multiple team and business unit support in logstore (#4041)
• feat: OpenAI compaction support (#4053)
• feat: multiple budget support for customers (#3998)
• feat: calendar_aligned budget support for customers (#3997)
• feat: provider-level governance for budgets & limits (#3938)
• feat: scope-aware budgets & limits wired from model configs (#3937, #3939, #3981)
• feat: created_by user attribution column for virtual keys (#3672)
• feat: blacklisted_models column for virtual key provider configs (#3653)
• feat: alias column in materialized view with filter support (#4078)
• feat: request_headers pattern capture for OTel and Maxim plugins (#4012)
• feat: paginated virtual key fetch for large key counts (#3957)
• feat: optional tx param on UpdateBudgetUsage for transaction support (#4039)
• feat: config.json source-of-truth flow (#3968)
• fix: reduce FTS index left() cap from 800k to 250k chars to stay within tsvector limit (#4057)
• fix: cascade deletes for model configs (#4051)
• fix: remove stale in-memory model configs (#4043)
• fix: add monotonic inc_number log cursor so node usage reconciliation does not skip late async log writes (#3664)
• fix: reduce sync worker ticker period to 5m to prevent threshold drift (#4023)
• fix: high-scale virtual key flow improvements (#4007)
• fix: preserve non-pricing model pool entries across pricing reloads (#3999)
• fix: passthrough budgets (#3941)
• refactor: make scope-level check methods extensible (#3940)
• revert: access_profile_id direct access profile assignment on virtual keys (#3669)
• chore: drop the access_profile_id column from governance_virtual_keys (#3670)

​

compat

0.1.16

• fix: preserve chat max_tokens during param filtering (#3992)
• chore: upgraded core to v1.5.17 and framework to v1.3.17

​

governance

1.5.17

• feat: provider-level governance for budgets & limits (#3938)
• feat: scope-aware budgets & limits wired from model configs (#3937, #3939, #3981)
• feat: team budget and rate-limit collection exporters (#4040)
• feat: multiple budget support for customers (#3998)
• feat: calendar_aligned budget support for customers (#3997)
• fix: replicate VK hierarchy flow for ghost node reconciliation (#4088)
• fix: cascade deletes for model configs (#4051)
• fix: create virtual key double usage counting (#4070)
• fix: high-scale virtual key flow improvements (#4007)
• fix: remove stale in-memory model configs (#4043)
• fix: gate models for passthrough in VK if present (#3988)
• fix: restrict fallbacks and provider selection to VK boundary (#3924)
• fix: passthrough budgets (#3941)
• refactor: make scope-level check methods extensible (#3940)

​

jsonparser

1.5.17

• chore: upgraded core to v1.5.17 and framework to v1.3.17

​

logging

1.5.17

• feat: multiple customers logs and usage tracking (#4066)
• feat: multiple team and business unit support in logstore (#4041)
• feat: OpenAI compaction support (#4053)
• feat: request_headers pattern capture with wildcard support (#4012)
• feat: wildcard pattern support for logging header capture, e.g. x-custom-* (#3958)
• feat: render file attachments in logs (#3931)
• fix: passthrough budgets (#3941)

​

maxim

1.6.17

• feat: request_headers pattern capture with wildcard support (#4012)

​

mocker

1.5.17

• chore: upgraded core to v1.5.17 and framework to v1.3.17

​

otel

1.2.17

• feat: disable_content_logging option to drop message/tool content from exported spans (#4064)
• feat: request_headers pattern capture with wildcard support (#4012)
• feat: multiple OTel collectors support (#3894)

​

prompts

1.0.17

• chore: upgraded core to v1.5.17 and framework to v1.3.17

​

semanticcache

1.5.17

• chore: upgraded core to v1.5.17 and framework to v1.3.17

​

telemetry

1.5.17

• chore: upgraded core to v1.5.17 and framework to v1.3.17